diff --git a/.gitignore b/.gitignore
index f14b8c6b78..e3fbd35171 100644
--- a/.gitignore
+++ b/.gitignore
@@ -56,7 +56,7 @@ cover/
*.pot
# Django stuff:
-*.log
+# *.log
local_settings.py
db.sqlite3
db.sqlite3-journal
@@ -208,3 +208,4 @@ __marimo__/
# LangGraph
.langgraph_api
+**/logs/agent_*.log
\ No newline at end of file
diff --git a/README.md b/README.md
index 7d3e938902..5fbc74ef5c 100644
--- a/README.md
+++ b/README.md
@@ -187,3 +187,15 @@ as well as custom instructions.
[] Create an example of a deep coding agent built on top of this
[] Benchmark the example of [deep research agent](examples/research/research_agent.py)
[] Add human-in-the-loop support for tools
+
+## Added InitChatModel [Ollama Qwen3](https://ollama.com/library/qwen3)
+
+Initial runs of out-of-box setup was took $1+, made mistake on first running `.invoke` instead of `.stream`
+
+```py
+init_chat_model(
+ model="ollama:qwen3:14b",
+ temperature=0.0,
+ max_tokens=40000,
+)
+```
\ No newline at end of file
diff --git a/examples/research/.vscode/launch.json b/examples/research/.vscode/launch.json
new file mode 100644
index 0000000000..1251d85c51
--- /dev/null
+++ b/examples/research/.vscode/launch.json
@@ -0,0 +1,31 @@
+{
+ "version": "0.2.0",
+ "configurations": [
+ {
+ "name": "Debug Research Agent",
+ "type": "python",
+ "request": "launch",
+ "program": "${workspaceFolder}/research_agent.py",
+ "console": "integratedTerminal",
+ "cwd": "${workspaceFolder}",
+ "env": {
+ "PYTHONPATH": "${workspaceFolder}"
+ },
+ "justMyCode": false,
+ "python": "${workspaceFolder}/.venv/bin/python"
+ },
+ {
+ "name": "Debug Research Agent (External Terminal)",
+ "type": "python",
+ "request": "launch",
+ "program": "${workspaceFolder}/research_agent.py",
+ "console": "externalTerminal",
+ "cwd": "${workspaceFolder}",
+ "env": {
+ "PYTHONPATH": "${workspaceFolder}"
+ },
+ "justMyCode": false,
+ "python": "${workspaceFolder}/.venv/bin/python"
+ }
+ ]
+}
\ No newline at end of file
diff --git a/examples/research/logs/ollama-deepagents-summary.log b/examples/research/logs/ollama-deepagents-summary.log
new file mode 100644
index 0000000000..12bdb607f2
--- /dev/null
+++ b/examples/research/logs/ollama-deepagents-summary.log
@@ -0,0 +1,166 @@
+2025-07-31 18:59:54 INFO: === Agent Message ===
+
+Okay, the user wants a summary of the GitHub repository https://github.com/hwchase17/deepagents. They mentioned reviewing the README, code, and social media. Let me break this down.
+
+First, I need to check the README. The user provided the URL, so I should use the internet_search function to look up the README. Wait, but the tools available include the internet_search function. However, the user specifically mentioned the README, code, and social media. Maybe I should first use the internet_search to get the README content. But the tools don't have a direct way to fetch the README. Alternatively, maybe the internet_search can be used with the query "README.md https://github.com/hwchase17/deepagents" to get the README content. But I'm not sure if that's the best approach.
+
+Wait, the tools available include the internet_search function. Let me check the parameters. The internet_search function requires a query, and can have a topic. The user wants the README, code, and social media. So perhaps I need to perform multiple internet searches. For the README, I can search for "README.md https://github.com/hwchase17/deepagents". For the code, maybe search for the repository's code structure. For social media, perhaps look for mentions on Twitter or other platforms.
+
+Alternatively, maybe the internet_search can be used with the query "summary of https://github.com/hwchase17/deepagents" to get a general overview. But the user specifically wants the README, code, and social media reviewed. So maybe three separate searches: one for the README, one for the code, and one for social media mentions.
+
+Wait, but the internet_search function's parameters include include_raw_content, max_results, query, and topic. The topic can be general, news, or finance. Maybe the social media part would be under general. But I'm not sure. Alternatively, maybe the code can be found by searching the repository's files. However, the tools don't have a direct way to list files or read them, except through internet_search.
+
+Another thought: the user might expect me to use the internet_search function to gather information about the repository. So first, I can search for the README content by querying "README.md https://github.com/hwchase17/deepagents" with include_raw_content set to true to get the actual content. Then, for the code, maybe search for key files or functions. For social media, search for mentions of the repository on platforms like Twitter.
+
+But the tools don't have a way to directly access the code files. The internet_search might return links to the code, but not the actual code content. Alternatively, maybe the README will have information about the code structure, and the social media links might be in the README or the repository's social media sections.
+
+So, the plan is:
+
+1. Use internet_search to get the README content by querying "README.md https://github.com/hwchase17/deepagents" with include_raw_content: true and max_results: 1 to get the main README.
+
+2. Use internet_search again to look for code-related information, maybe by searching for "code https://github.com/hwchase17/deepagents" to find code files or structure.
+
+3. Use internet_search to find social media mentions by querying "social media https://github.com/hwchase17/deepagents" or similar.
+
+But I need to make sure that the queries are specific enough. Alternatively, maybe the first search for the README will include links to the code and social media sections. However, the user explicitly mentioned reviewing the README, code, and social media, so separate searches might be necessary.
+
+Another consideration: the internet_search function's topic parameter can be set to "general" for general information, but the user might need specific details. However, the tools don't have a way to directly access the code or social media links except through web search.
+
+So, the first step is to perform an internet search for the README content. Then, another search for code details, and another for social media mentions. Each search would be a separate function call.
+
+2025-07-31 18:59:56 INFO: +++ Tool Call +++
+{"query": "code structure https://github.com/hwchase17/deepagents", "follow_up_questions": null, "answer": null, "images": [], "results": [{"url": "https://github.com/hwchase17/deepagents", "title": "hwchase17/deepagents - GitHub", "content": "deepagents is a Python package that implements these in a general purpose way so that you can easily create a Deep Agent for your application.", "score": 0.75723875, "raw_content": "## Navigation Menu\n\n# Search code, repositories, users, issues, pull requests...\n\n# Provide feedback\n\nWe read every piece of feedback, and take your input very seriously.\n\n# Saved searches\n\n## Use saved searches to filter your results more quickly\n\nTo see all available qualifiers, see our [documentation](https://docs.github.com/search-github/github-code-search/understanding-github-code-search-syntax).\n\n### License\n\n### Uh oh!\n\nThere was an error while loading. Please reload this page.\n\nThere was an error while loading. Please reload this page.\n\n# hwchase17/deepagents\n\n## Folders and files\n\n| Name | | Name | Last commit message | Last commit date |\n| --- | --- | --- | --- | --- |\n| Latest commit History[8 Commits](/hwchase17/deepagents/commits/master/) | | |\n| [examples/research](/hwchase17/deepagents/tree/master/examples/research \"This path skips through empty directories\") | | [examples/research](/hwchase17/deepagents/tree/master/examples/research \"This path skips through empty directories\") | | |\n| [src/deepagents](/hwchase17/deepagents/tree/master/src/deepagents \"This path skips through empty directories\") | | [src/deepagents](/hwchase17/deepagents/tree/master/src/deepagents \"This path skips through empty directories\") | | |\n| [.gitignore](/hwchase17/deepagents/blob/master/.gitignore \".gitignore\") | | [.gitignore](/hwchase17/deepagents/blob/master/.gitignore \".gitignore\") | | |\n| [LICENSE](/hwchase17/deepagents/blob/master/LICENSE \"LICENSE\") | | [LICENSE](/hwchase17/deepagents/blob/master/LICENSE \"LICENSE\") | | |\n| [README.md](/hwchase17/deepagents/blob/master/README.md \"README.md\") | | [README.md](/hwchase17/deepagents/blob/master/README.md \"README.md\") | | |\n| [deep\\_agents.png](/hwchase17/deepagents/blob/master/deep_agents.png \"deep_agents.png\") | | [deep\\_agents.png](/hwchase17/deepagents/blob/master/deep_agents.png \"deep_agents.png\") | | |\n| [pyproject.toml](/hwchase17/deepagents/blob/master/pyproject.toml \"pyproject.toml\") | | [pyproject.toml](/hwchase17/deepagents/blob/master/pyproject.toml \"pyproject.toml\") | | |\n| View all files | | |\n\n## Latest commit\n\n## History\n\n## Repository files navigation\n\n# π§ π€Deep Agents\n\nUsing an LLM to call tools in a loop is the simplest form of an agent.\nThis architecture, however, can yield agents that are βshallowβ and fail to plan and act over longer, more complex tasks.\nApplications like βDeep Researchβ, \"Manus\", and βClaude Codeβ have gotten around this limitation by implementing a combination of four things:\na **planning tool**, **sub agents**, access to a **file system**, and a **detailed prompt**.\n\n[](/hwchase17/deepagents/blob/master/deep_agents.png)\n\n\n\n`deepagents` is a Python package that implements these in a general purpose way so that you can easily create a Deep Agent for your application.\n\n`deepagents`\n\n**Acknowledgements: This project was primarily inspired by Claude Code, and initially was largely an attempt to see what made Claude Code general purpose, and make it even more so.**\n\n## Installation\n\n## Usage\n\n(To run the example below, will need to `pip install tavily-python`)\n\n`pip install tavily-python`\n\nSee [examples/research/research\\_agent.py](/hwchase17/deepagents/blob/master/examples/research/research_agent.py) for a more complex example.\n\nThe agent created with `create_deep_agent` is just a LangGraph graph - so you can interact with it (streaming, human-in-the-loop, memory, studio)\nin the same way you would any LangGraph agent.\n\n`create_deep_agent`\n\n## Creating a custom deep agent\n\nThere are three parameters you can pass to `create_deep_agent` to create your own custom deep agent.\n\n`create_deep_agent`\n\n### `tools` (Required)\n\n`tools`\n\nThe first argument to `create_deep_agent` is `tools`.\nThis should be a list of functions or LangChain `@tool` objects.\nThe agent (and any subagents) will have access to these tools.\n\n`create_deep_agent`\n`tools`\n`@tool`\n\n### `instructions` (Required)\n\n`instructions`\n\nThe second argument to `create_deep_agent` is `instructions`.\nThis will serve as part of the prompt of the deep agent.\nNote that there is a [built in system prompt](#built-in-prompt) as well, so this is not the *entire* prompt the agent will see.\n\n`create_deep_agent`\n`instructions`\n\n### `subagents` (Optional)\n\n`subagents`\n\nA keyword-only argument to `create_deep_agent` is `subagents`.\nThis can be used to specify any custom subagents this deep agent will have access to.\nYou can read more about why you would want to use subagents [here](#sub-agents)\n\n`create_deep_agent`\n`subagents`\n\n`subagents` should be a list of dictionaries, where each dictionary follow this schema:\n\n`subagents`\n\nTo use it looks like:\n\n### `model` (Optional)\n\n`model`\n\nBy default, `deepagents` will use `\"claude-sonnet-4-20250514\"`. If you want to use a different model,\nyou can pass a [LangChain model object](https://python.langchain.com/docs/integrations/chat/).\n\n`deepagents`\n`\"claude-sonnet-4-20250514\"`\n\n## Deep Agent Details\n\nThe below components are built into `deepagents` and helps make it work for deep tasks off-the-shelf.\n\n`deepagents`\n\n### System Prompt\n\n`deepagents` comes with a [built-in system prompt](/hwchase17/deepagents/blob/master/src/deepagents/prompts.py). This is relatively detailed prompt that is heavily based on and inspired by [attempts](https://github.com/kn1026/cc/blob/main/claudecode.md) to [replicate](https://github.com/asgeirtj/system_prompts_leaks/blob/main/Anthropic/claude-code.md)\nClaude Code's system prompt. It was made more general purpose than Claude Code's system prompt.\nThis contains detailed instructions for how to use the built-in planning tool, file system tools, and sub agents.\nNote that part of this system prompt [can be customized](#promptprefix--required-)\n\n`deepagents`\n\nWithout this default system prompt - the agent would not be nearly as successful at going as it is.\nThe importance of prompting for creating a \"deep\" agent cannot be understated.\n\n### Planing Tool\n\n`deepagents` comes with a built-in planning tool. This planning tool is very simple and is based on ClaudeCode's TodoWrite tool.\nThis tool doesn't actually do anything - it is just a way for the agent to come up with a plan, and then have that in the context to help keep it on track.\n\n`deepagents`\n\n### File System Tools\n\n`deepagents` comes with four built-in file system tools: `ls`, `edit_file`, `read_file`, `write_file`.\nThese do not actually use a file system - rather, they mock out a file system using LangGraph's State object.\nThis means you can easily run many of these agents on the same machine without worrying that they will edit the same underlying files.\n\n`deepagents`\n`ls`\n`edit_file`\n`read_file`\n`write_file`\n\nRight now the \"file system\" will only be one level deep (no sub directories).\n\nThese files can be passed in (and also retrieved) by using the `files` key in the LangGraph State object.\n\n`files`\n\n### Sub Agents\n\n`deepagents` comes with the built-in ability to call sub agents (based on Claude Code).\nIt has access to a `general-purpose` subagent at all times - this is a subagent with the same instructions as the main agent and all the tools that is has access to.\nYou can also specify [custom sub agents](#subagents--optional-) with their own instructions and tools.\n\n`deepagents`\n`general-purpose`\n\nSub agents are useful for [\"context quarantine\"](https://www.dbreunig.com/2025/06/26/how-to-fix-your-context.html#context-quarantine) (to help not pollute the overall context of the main agent)\nas well as custom instructions.\n\n## Roadmap\n\n[] Allow users to customize full system prompt\n[] Code cleanliness (type hinting, docstrings, formating)\n[] Allow for more of a robust virtual filesystem\n[] Create an example of a deep coding agent built on top of this\n[] Benchmark the example of [deep research agent](/hwchase17/deepagents/blob/master/examples/research/research_agent.py)\n[] Add human-in-the-loop support for tools\n\n## About\n\n### Resources\n\n### License\n\n### Uh oh!\n\nThere was an error while loading. Please reload this page.\n\nThere was an error while loading. Please reload this page.\n\n### Stars\n\n### Watchers\n\n### Forks\n\n## [Releases](/hwchase17/deepagents/releases)\n\n## [Packages 0](/users/hwchase17/packages?repo_name=deepagents)\n\n## Languages\n\n## Footer\n\n### Footer navigation"}, {"url": "https://github.com/langchain-ai/langchain", "title": "langchain-ai/langchain: Build context-aware reasoning applications", "content": "### Dependencies none ### Tag maintainer @baskaryan ### Twitter handle kzk_maeda\")#11259) ### Description renamed several repository links from `hwchase17` to `langchain-ai`. -Description: Adding new package: `langchain-box`: * `langchain_box.document_loaders.BoxLoader` β DocumentLoader functionality * `langchain_box.utilities.BoxAPIWrapper` β Box-specific code * `langchain_box.utilities.BoxAuth` β Helper class for Box authentication * `langchain_box.utilities.BoxAuthType` β enum used by BoxAuth class - Twitter handle: @boxplatform - [x] **Add tests and docs**: If you're adding a new integration, please include 1. -Description: Adding new package: `langchain-box`: * `langchain_box.document_loaders.BoxLoader` β DocumentLoader functionality * `langchain_box.utilities.BoxAPIWrapper` β Box-specific code * `langchain_box.utilities.BoxAuth` β Helper class for Box authentication * `langchain_box.utilities.BoxAuthType` β enum used by BoxAuth class - Twitter handle: @boxplatform - [x] **Add tests and docs**: If you're adding a new integration, please include 1.", "score": 0.23472267, "raw_content": "GitHub - langchain-ai/langchain: π¦π Build context-aware reasoning applications\n\n\n\n[Skip to content](#start-of-content)\n\n\n\n\n\n\n\n\n\n\n\n\n## Navigation Menu\n\nToggle navigation\n\n[Sign in](/login?return_to=https%3A%2F%2Fgithub.com%2Flangchain-ai%2Flangchain)\n\nAppearance settings\n\n- Product\n\n - [GitHub Copilot\n\n Write better code with AI](https://github.com/features/copilot)\n - [GitHub Models\n New\n\n Manage and compare prompts](https://github.com/features/models)\n - [GitHub Advanced Security\n\n Find and fix vulnerabilities](https://github.com/security/advanced-security)\n - [Actions\n\n Automate any workflow](https://github.com/features/actions)\n - [Codespaces\n\n Instant dev environments](https://github.com/features/codespaces)\n\n - [Issues\n\n Plan and track work](https://github.com/features/issues)\n - [Code Review\n\n Manage code changes](https://github.com/features/code-review)\n - [Discussions\n\n Collaborate outside of code](https://github.com/features/discussions)\n - [Code Search\n\n Find more, search less](https://github.com/features/code-search)\n\n Explore\n - [Why GitHub](https://github.com/why-github)\n - [All features](https://github.com/features)\n - [Documentation](https://docs.github.com)\n - [GitHub Skills](https://skills.github.com)\n - [Blog](https://github.blog)\n- Solutions\n\n By company size\n - [Enterprises](https://github.com/enterprise)\n - [Small and medium teams](https://github.com/team)\n - [Startups](https://github.com/enterprise/startups)\n - [Nonprofits](/solutions/industry/nonprofits)\n\n By use case\n - [DevSecOps](/solutions/use-case/devsecops)\n - [DevOps](/solutions/use-case/devops)\n - [CI/CD](/solutions/use-case/ci-cd)\n - [View all use cases](/solutions/use-case)\n\n By industry\n - [Healthcare](/solutions/industry/healthcare)\n - [Financial services](/solutions/industry/financial-services)\n - [Manufacturing](/solutions/industry/manufacturing)\n - [Government](/solutions/industry/government)\n - [View all industries](/solutions/industry)\n\n [View all solutions](/solutions)\n- Resources\n\n Topics\n - [AI](/resources/articles/ai)\n - [DevOps](/resources/articles/devops)\n - [Security](/resources/articles/security)\n - [Software Development](/resources/articles/software-development)\n - [View all](/resources/articles)\n\n Explore\n - [Learning Pathways](https://resources.github.com/learn/pathways)\n - [Events & Webinars](https://resources.github.com)\n - [Ebooks & Whitepapers](https://github.com/resources/whitepapers)\n - [Customer Stories](https://github.com/customer-stories)\n - [Partners](https://partner.github.com)\n - [Executive Insights](https://github.com/solutions/executive-insights)\n- Open Source\n\n - [GitHub Sponsors\n\n Fund open source developers](/sponsors)\n\n - [The ReadME Project\n\n GitHub community articles](https://github.com/readme)\n\n Repositories\n - [Topics](https://github.com/topics)\n - [Trending](https://github.com/trending)\n - [Collections](https://github.com/collections)\n- Enterprise\n\n - [Enterprise platform\n\n AI-powered developer platform](/enterprise)\n\n Available add-ons\n - [GitHub Advanced Security\n\n Enterprise-grade security features](https://github.com/security/advanced-security)\n - [Copilot for business\n\n Enterprise-grade AI features](/features/copilot/copilot-business)\n - [Premium Support\n\n Enterprise-grade 24/7 support](/premium-support)\n- [Pricing](https://github.com/pricing)\n\nSearch or jump to...\n\n\n# Search code, repositories, users, issues, pull requests...\n\nSearch\n\nClear\n\n[Search syntax tips](https://docs.github.com/search-github/github-code-search/understanding-github-code-search-syntax)\n\n# Provide feedback\n\nWe read every piece of feedback, and take your input very seriously.\n\n\nInclude my email address so I can be contacted\n\nCancel\n Submit feedback\n\n\n\n\n\n# Saved searches\n\n## Use saved searches to filter your results more quickly\n\nName\n\nQuery\n\nTo see all available qualifiers, see our [documentation](https://docs.github.com/search-github/github-code-search/understanding-github-code-search-syntax).\n\nCancel\n Create saved search\n\n[Sign in](/login?return_to=https%3A%2F%2Fgithub.com%2Flangchain-ai%2Flangchain)\n\n[Sign up](/signup?ref_cta=Sign+up&ref_loc=header+logged+out&ref_page=%2F%3Cuser-name%3E%2F%3Crepo-name%3E&source=header-repo&source_repo=langchain-ai%2Flangchain)\n\nAppearance settings\n\nResetting focus\n\nYou signed in with another tab or window. Reload to refresh your session.\nYou signed out in another tab or window. Reload to refresh your session.\nYou switched accounts on another tab or window. Reload to refresh your session.\n \n\n\nDismiss alert\n\n{{ message }}\n\n[langchain-ai](/langchain-ai) \n/\n**[langchain](/langchain-ai/langchain)**\nPublic\n\n- [Notifications](/login?return_to=%2Flangchain-ai%2Flangchain) You must be signed in to change notification settings\n- [Fork\n 17.8k](/login?return_to=%2Flangchain-ai%2Flangchain)\n- [Star\n 109k](/login?return_to=%2Flangchain-ai%2Flangchain)\n\nπ¦π Build context-aware reasoning applications\n\n[python.langchain.com](https://python.langchain.com \"https://python.langchain.com\")\n\n### License\n\n[MIT license](/langchain-ai/langchain/blob/master/LICENSE)\n\n[109k\nstars](/langchain-ai/langchain/stargazers) [17.8k\nforks](/langchain-ai/langchain/forks) [Branches](/langchain-ai/langchain/branches) [Tags](/langchain-ai/langchain/tags) [Activity](/langchain-ai/langchain/activity)\n\n[Star](/login?return_to=%2Flangchain-ai%2Flangchain)\n\n[Notifications](/login?return_to=%2Flangchain-ai%2Flangchain) You must be signed in to change notification settings\n\n- [Code](/langchain-ai/langchain)\n- [Issues\n 246](/langchain-ai/langchain/issues)\n- [Pull requests\n 54](/langchain-ai/langchain/pulls)\n- [Discussions](/langchain-ai/langchain/discussions)\n- [Actions](/langchain-ai/langchain/actions)\n- [Projects\n 2](/langchain-ai/langchain/projects)\n- [Security](/langchain-ai/langchain/security)\n\n [### Uh oh!](/langchain-ai/langchain/security)\n\n [There was an error while loading.](/langchain-ai/langchain/security) Please reload this page.\n- [Insights](/langchain-ai/langchain/pulse)\n\nAdditional navigation options\n\n\n- [Code](/langchain-ai/langchain)\n- [Issues](/langchain-ai/langchain/issues)\n- [Pull requests](/langchain-ai/langchain/pulls)\n- [Discussions](/langchain-ai/langchain/discussions)\n- [Actions](/langchain-ai/langchain/actions)\n- [Projects](/langchain-ai/langchain/projects)\n- [Security](/langchain-ai/langchain/security)\n- [Insights](/langchain-ai/langchain/pulse)\n\n# langchain-ai/langchain\n\nmaster\n\n[**901** Branches](/langchain-ai/langchain/branches)[**906** Tags](/langchain-ai/langchain/tags)\n\nGo to file\n\nCode\n\nOpen more actions menu\n\n## Folders and files\n\n| Name | | Name | Last commit message | Last commit date |\n| --- | --- | --- | --- | --- |\n| Latest commit [ccurme](/ccurme)[ccurme](/langchain-ai/langchain/commits?author=ccurme) [docs: document built-in tools for ChatVertexAI (](/langchain-ai/langchain/commit/446a9d5647918f1dd6b979122bf6f08e50b8a79e)[#31564](https://github.com/langchain-ai/langchain/pull/31564)[)](/langchain-ai/langchain/commit/446a9d5647918f1dd6b979122bf6f08e50b8a79e) Jun 11, 2025 [446a9d5](/langchain-ai/langchain/commit/446a9d5647918f1dd6b979122bf6f08e50b8a79e)Β Β·Β Jun 11, 2025 History[13,531 Commits](/langchain-ai/langchain/commits/master/) Open commit details | | |\n| [.devcontainer](/langchain-ai/langchain/tree/master/.devcontainer \".devcontainer\") | | [.devcontainer](/langchain-ai/langchain/tree/master/.devcontainer \".devcontainer\") | [community[minor]: Add ApertureDB as a vectorstore (](/langchain-ai/langchain/commit/5caa3811770fe25e54e46b61e1b14d1a985a0a32 \"community[minor]: Add ApertureDB as a vectorstore (#24088) Thank you for contributing to LangChain! - [X] *ApertureDB as vectorstore**: \\\"community: Add ApertureDB as a vectorestore\\\" - **Description:** this change provides a new community integration that uses ApertureData's ApertureDB as a vector store. - **Issue:** none - **Dependencies:** depends on ApertureDB Python SDK - **Twitter handle:** ApertureData - [X] **Add tests and docs**: If you're adding a new integration, please include 1. a test for the integration, preferably unit tests that do not rely on network access, 2. an example notebook showing its use. It lives in `docs/docs/integrations` directory. Integration tests rely on a local run of a public docker image. Example notebook additionally relies on a local Ollama server. - [X] **Lint and test**: Run `make format`, `make lint` and `make test` from the root of the package(s) you've modified. See contribution guidelines for more: https://python.langchain.com/docs/contributing/ All lint tests pass. Additional guidelines: - Make sure optional dependencies are imported within a function. - Please do not add dependencies to pyproject.toml files (even optional ones) unless they are required for unit tests. - Most PRs should not touch more than one package. - Changes should be backwards compatible. - If you are adding something to community, do not re-import it in langchain. If no one reviews your PR within a few days, please @-mention one of baskaryan, efriis, eyurtsev, ccurme, vbarda, hwchase17. --------- Co-authored-by: Gautam \")[#24088](https://github.com/langchain-ai/langchain/pull/24088)[)](/langchain-ai/langchain/commit/5caa3811770fe25e54e46b61e1b14d1a985a0a32 \"community[minor]: Add ApertureDB as a vectorstore (#24088) Thank you for contributing to LangChain! - [X] *ApertureDB as vectorstore**: \\\"community: Add ApertureDB as a vectorestore\\\" - **Description:** this change provides a new community integration that uses ApertureData's ApertureDB as a vector store. - **Issue:** none - **Dependencies:** depends on ApertureDB Python SDK - **Twitter handle:** ApertureData - [X] **Add tests and docs**: If you're adding a new integration, please include 1. a test for the integration, preferably unit tests that do not rely on network access, 2. an example notebook showing its use. It lives in `docs/docs/integrations` directory. Integration tests rely on a local run of a public docker image. Example notebook additionally relies on a local Ollama server. - [X] **Lint and test**: Run `make format`, `make lint` and `make test` from the root of the package(s) you've modified. See contribution guidelines for more: https://python.langchain.com/docs/contributing/ All lint tests pass. Additional guidelines: - Make sure optional dependencies are imported within a function. - Please do not add dependencies to pyproject.toml files (even optional ones) unless they are required for unit tests. - Most PRs should not touch more than one package. - Changes should be backwards compatible. - If you are adding something to community, do not re-import it in langchain. If no one reviews your PR within a few days, please @-mention one of baskaryan, efriis, eyurtsev, ccurme, vbarda, hwchase17. --------- Co-authored-by: Gautam \") | Jul 16, 2024 |\n| [.github](/langchain-ai/langchain/tree/master/.github \".github\") | | [.github](/langchain-ai/langchain/tree/master/.github \".github\") | [infra: drop azure from streaming benchmarks (](/langchain-ai/langchain/commit/c8951ca124ca84d922f94c9336bb82e7d83a6d28 \"infra: drop azure from streaming benchmarks (#31421) Covered by BaseChatOpenAI\")[#31421](https://github.com/langchain-ai/langchain/pull/31421)[)](/langchain-ai/langchain/commit/c8951ca124ca84d922f94c9336bb82e7d83a6d28 \"infra: drop azure from streaming benchmarks (#31421) Covered by BaseChatOpenAI\") | May 29, 2025 |\n| [cookbook](/langchain-ai/langchain/tree/master/cookbook \"cookbook\") | | [cookbook](/langchain-ai/langchain/tree/master/cookbook \"cookbook\") | [docs: Update Multi\\_modal\\_RAG\\_google.ipynb to remove the unsupported Gβ¦](/langchain-ai/langchain/commit/e455fab5d3ca2820db2f49c5511a23b168a441ab \"docs: Update Multi_modal_RAG_google.ipynb to remove the unsupported Gemini models (#31526) - **Description:** Remove the outdated Gemini models and replace those with the latest models. - **Issue:** Earlier the code was not running, now the code runs. - **Dependencies:** No - **Twitter handle:** [soumendrak_](https://x.com/soumendrak_)\") | Jun 9, 2025 |\n| [docs](/langchain-ai/langchain/tree/master/docs \"docs\") | | [docs](/langchain-ai/langchain/tree/master/docs \"docs\") | [docs: document built-in tools for ChatVertexAI (](/langchain-ai/langchain/commit/446a9d5647918f1dd6b979122bf6f08e50b8a79e \"docs: document built-in tools for ChatVertexAI (#31564)\")[#31564](https://github.com/langchain-ai/langchain/pull/31564)[)](/langchain-ai/langchain/commit/446a9d5647918f1dd6b979122bf6f08e50b8a79e \"docs: document built-in tools for ChatVertexAI (#31564)\") | Jun 11, 2025 |\n| [libs](/langchain-ai/langchain/tree/master/libs \"libs\") | | [libs](/langchain-ai/langchain/tree/master/libs \"libs\") | [langchain[patch]: Allow specifying other hashing functions in embeddiβ¦](/langchain-ai/langchain/commit/d10fd02bb33d36d43d86ec6a3c7489b15edab8c8 \"langchain[patch]: Allow specifying other hashing functions in embeddings (#31561) Allow specifying other hashing functions in embeddings\") | Jun 11, 2025 |\n| [scripts](/langchain-ai/langchain/tree/master/scripts \"scripts\") | | [scripts](/langchain-ai/langchain/tree/master/scripts \"scripts\") | [infra: update mypy 1.10, ruff 0.5 (](/langchain-ai/langchain/commit/a0c2281540885b068331921d45d2549f7df8b55f \"infra: update mypy 1.10, ruff 0.5 (#23721) ```python \\\"\\\"\\\"python scripts/update_mypy_ruff.py\\\"\\\"\\\" import glob import tomllib from pathlib import Path import toml import subprocess import re ROOT_DIR = Path(__file__).parents[1] def main(): for path in glob.glob(str(ROOT_DIR / \\\"libs/**/pyproject.toml\\\"), recursive=True): print(path) with open(path, \\\"rb\\\") as f: pyproject = tomllib.load(f) try: pyproject[\\\"tool\\\"][\\\"poetry\\\"][\\\"group\\\"][\\\"typing\\\"][\\\"dependencies\\\"][\\\"mypy\\\"] = ( \\\"^1.10\\\" ) pyproject[\\\"tool\\\"][\\\"poetry\\\"][\\\"group\\\"][\\\"lint\\\"][\\\"dependencies\\\"][\\\"ruff\\\"] = ( \\\"^0.5\\\" ) except KeyError: continue with open(path, \\\"w\\\") as f: toml.dump(pyproject, f) cwd = \\\"/\\\".join(path.split(\\\"/\\\")[:-1]) completed = subprocess.run( \\\"poetry lock --no-update; poetry install --with typing; poetry run mypy . --no-color\\\", cwd=cwd, shell=True, capture_output=True, text=True, ) logs = completed.stdout.split(\\\"\\n\\\") to_ignore = {} for l in logs: if re.match(\\\"^(.*)\\:(\\d+)\\: error:.*\\[(.*)\\]\\\", l): path, line_no, error_type = re.match( \\\"^(.*)\\:(\\d+)\\: error:.*\\[(.*)\\]\\\", l ).groups() if (path, line_no) in to_ignore: to_ignore[(path, line_no)].append(error_type) else: to_ignore[(path, line_no)] = [error_type] print(len(to_ignore)) for (error_path, line_no), error_types in to_ignore.items(): all_errors = \\\", \\\".join(error_types) full_path = f\\\"{cwd}/{error_path}\\\" try: with open(full_path, \\\"r\\\") as f: file_lines = f.readlines() except FileNotFoundError: continue file_lines[int(line_no) - 1] = ( file_lines[int(line_no) - 1][:-1] + f\\\" # type: ignore[{all_errors}]\\n\\\" ) with open(full_path, \\\"w\\\") as f: f.write(\\\"\\\".join(file_lines)) subprocess.run( \\\"poetry run ruff format .; poetry run ruff --select I --fix .\\\", cwd=cwd, shell=True, capture_output=True, text=True, ) if __name__ == \\\"__main__\\\": main() ```\")[#23721](https://github.com/langchain-ai/langchain/pull/23721)[)](/langchain-ai/langchain/commit/a0c2281540885b068331921d45d2549f7df8b55f \"infra: update mypy 1.10, ruff 0.5 (#23721) ```python \\\"\\\"\\\"python scripts/update_mypy_ruff.py\\\"\\\"\\\" import glob import tomllib from pathlib import Path import toml import subprocess import re ROOT_DIR = Path(__file__).parents[1] def main(): for path in glob.glob(str(ROOT_DIR / \\\"libs/**/pyproject.toml\\\"), recursive=True): print(path) with open(path, \\\"rb\\\") as f: pyproject = tomllib.load(f) try: pyproject[\\\"tool\\\"][\\\"poetry\\\"][\\\"group\\\"][\\\"typing\\\"][\\\"dependencies\\\"][\\\"mypy\\\"] = ( \\\"^1.10\\\" ) pyproject[\\\"tool\\\"][\\\"poetry\\\"][\\\"group\\\"][\\\"lint\\\"][\\\"dependencies\\\"][\\\"ruff\\\"] = ( \\\"^0.5\\\" ) except KeyError: continue with open(path, \\\"w\\\") as f: toml.dump(pyproject, f) cwd = \\\"/\\\".join(path.split(\\\"/\\\")[:-1]) completed = subprocess.run( \\\"poetry lock --no-update; poetry install --with typing; poetry run mypy . --no-color\\\", cwd=cwd, shell=True, capture_output=True, text=True, ) logs = completed.stdout.split(\\\"\\n\\\") to_ignore = {} for l in logs: if re.match(\\\"^(.*)\\:(\\d+)\\: error:.*\\[(.*)\\]\\\", l): path, line_no, error_type = re.match( \\\"^(.*)\\:(\\d+)\\: error:.*\\[(.*)\\]\\\", l ).groups() if (path, line_no) in to_ignore: to_ignore[(path, line_no)].append(error_type) else: to_ignore[(path, line_no)] = [error_type] print(len(to_ignore)) for (error_path, line_no), error_types in to_ignore.items(): all_errors = \\\", \\\".join(error_types) full_path = f\\\"{cwd}/{error_path}\\\" try: with open(full_path, \\\"r\\\") as f: file_lines = f.readlines() except FileNotFoundError: continue file_lines[int(line_no) - 1] = ( file_lines[int(line_no) - 1][:-1] + f\\\" # type: ignore[{all_errors}]\\n\\\" ) with open(full_path, \\\"w\\\") as f: f.write(\\\"\\\".join(file_lines)) subprocess.run( \\\"poetry run ruff format .; poetry run ruff --select I --fix .\\\", cwd=cwd, shell=True, capture_output=True, text=True, ) if __name__ == \\\"__main__\\\": main() ```\") | Jul 3, 2024 |\n| [.gitattributes](/langchain-ai/langchain/blob/master/.gitattributes \".gitattributes\") | | [.gitattributes](/langchain-ai/langchain/blob/master/.gitattributes \".gitattributes\") | [Update dev container (](/langchain-ai/langchain/commit/ccd916babe8b24ff8a2ef5d9c2a8a53487388f2a \"Update dev container (#6189) Fixes https://github.com/hwchase17/langchain/issues/6172 As described in https://github.com/hwchase17/langchain/issues/6172, I'd love to help update the dev container in this project. **Summary of changes:** - Dev container now builds (the current container in this repo won't build for me) - Dockerfile updates - Update image to our [currently-maintained Python image](https://github.com/devcontainers/images/tree/main/src/python/.devcontainer) (`mcr.microsoft.com/devcontainers/python`) rather than the deprecated image from vscode-dev-containers - Move Dockerfile to root of repo - in order for `COPY` to work properly, it needs the files (in this case, `pyproject.toml` and `poetry.toml`) in the same directory - devcontainer.json updates - Removed `customizations` and `remoteUser` since they should be covered by the updated image in the Dockerfile - Update comments - Update docker-compose.yaml to properly point to updated Dockerfile - Add a .gitattributes to avoid line ending conversions, which can result in hundreds of pending changes ([info](https://code.visualstudio.com/docs/devcontainers/tips-and-tricks#_resolving-git-line-ending-issues-in-containers-resulting-in-many-modified-files)) - Add a README in the .devcontainer folder and info on the dev container in the contributing.md **Outstanding questions:** - Is it expected for `poetry install` to take some time? It takes about 30 minutes for this dev container to finish building in a Codespace, but a user should only have to experience this once. Through some online investigation, this doesn't seem unusual - Versions of poetry newer than 1.3.2 failed every time - based on some of the guidance in contributing.md and other online resources, it seemed changing poetry versions might be a good solution. 1.3.2 is from Jan 2023 --------- Co-authored-by: bamurtaugh Co-authored-by: Samruddhi Khandale \")[#6189](https://github.com/langchain-ai/langchain/pull/6189)[)](/langchain-ai/langchain/commit/ccd916babe8b24ff8a2ef5d9c2a8a53487388f2a \"Update dev container (#6189) Fixes https://github.com/hwchase17/langchain/issues/6172 As described in https://github.com/hwchase17/langchain/issues/6172, I'd love to help update the dev container in this project. **Summary of changes:** - Dev container now builds (the current container in this repo won't build for me) - Dockerfile updates - Update image to our [currently-maintained Python image](https://github.com/devcontainers/images/tree/main/src/python/.devcontainer) (`mcr.microsoft.com/devcontainers/python`) rather than the deprecated image from vscode-dev-containers - Move Dockerfile to root of repo - in order for `COPY` to work properly, it needs the files (in this case, `pyproject.toml` and `poetry.toml`) in the same directory - devcontainer.json updates - Removed `customizations` and `remoteUser` since they should be covered by the updated image in the Dockerfile - Update comments - Update docker-compose.yaml to properly point to updated Dockerfile - Add a .gitattributes to avoid line ending conversions, which can result in hundreds of pending changes ([info](https://code.visualstudio.com/docs/devcontainers/tips-and-tricks#_resolving-git-line-ending-issues-in-containers-resulting-in-many-modified-files)) - Add a README in the .devcontainer folder and info on the dev container in the contributing.md **Outstanding questions:** - Is it expected for `poetry install` to take some time? It takes about 30 minutes for this dev container to finish building in a Codespace, but a user should only have to experience this once. Through some online investigation, this doesn't seem unusual - Versions of poetry newer than 1.3.2 failed every time - based on some of the guidance in contributing.md and other online resources, it seemed changing poetry versions might be a good solution. 1.3.2 is from Jan 2023 --------- Co-authored-by: bamurtaugh Co-authored-by: Samruddhi Khandale \") | Jun 16, 2023 |\n| [.gitignore](/langchain-ai/langchain/blob/master/.gitignore \".gitignore\") | | [.gitignore](/langchain-ai/langchain/blob/master/.gitignore \".gitignore\") | [[performance]: Adding benchmarks for common](/langchain-ai/langchain/commit/78ec7d886d8ff1337174e775df29f9553bcf859b \"[performance]: Adding benchmarks for common `langchain-core` imports (#30747) The first in a sequence of PRs focusing on improving performance in core. We're starting with reducing import times for common structures, hence the benchmarks here. The benchmark looks a little bit complicated - we have to use a process so that we don't suffer from Python's import caching system. I tried doing manual modification of `sys.modules` between runs, but that's pretty tricky / hacky to get right, hence the subprocess approach. Motivated by extremely slow baseline for common imports (we're talking 2-5 seconds): ![\\\"Screenshot](\\\"https://github.com/user-attachments/assets/994616fe-1798-404d-bcbe-48ad0eb8a9a0\\\")
Also added a `make benchmark` command to make local runs easy :). Currently using walltimes so that we can track total time despite using a manual proces.\") `langchain-core` [imports (](/langchain-ai/langchain/commit/78ec7d886d8ff1337174e775df29f9553bcf859b \"[performance]: Adding benchmarks for common `langchain-core` imports (#30747) The first in a sequence of PRs focusing on improving performance in core. We're starting with reducing import times for common structures, hence the benchmarks here. The benchmark looks a little bit complicated - we have to use a process so that we don't suffer from Python's import caching system. I tried doing manual modification of `sys.modules` between runs, but that's pretty tricky / hacky to get right, hence the subprocess approach. Motivated by extremely slow baseline for common imports (we're talking 2-5 seconds): Also added a `make benchmark` command to make local runs easy :). Currently using walltimes so that we can track total time despite using a manual proces.\")[#β¦](https://github.com/langchain-ai/langchain/pull/30747) | Apr 9, 2025 |\n| [.pre-commit-config.yaml](/langchain-ai/langchain/blob/master/.pre-commit-config.yaml \".pre-commit-config.yaml\") | | [.pre-commit-config.yaml](/langchain-ai/langchain/blob/master/.pre-commit-config.yaml \".pre-commit-config.yaml\") | [voyageai: remove from monorepo (](/langchain-ai/langchain/commit/bf645c83f4e58a5e09429d00b02ee73c6bf65270 \"voyageai: remove from monorepo (#31281) langchain-voyageai is now maintained at https://github.com/voyage-ai/langchain-voyageai.\")[#31281](https://github.com/langchain-ai/langchain/pull/31281)[)](/langchain-ai/langchain/commit/bf645c83f4e58a5e09429d00b02ee73c6bf65270 \"voyageai: remove from monorepo (#31281) langchain-voyageai is now maintained at https://github.com/voyage-ai/langchain-voyageai.\") | May 19, 2025 |\n| [.readthedocs.yaml](/langchain-ai/langchain/blob/master/.readthedocs.yaml \".readthedocs.yaml\") | | [.readthedocs.yaml](/langchain-ai/langchain/blob/master/.readthedocs.yaml \".readthedocs.yaml\") | [docs(readthedocs): streamline config (](/langchain-ai/langchain/commit/492b4c1604300385f5fe539876f9dab7f26ab870 \"docs(readthedocs): streamline config (#30307)\")[#30307](https://github.com/langchain-ai/langchain/pull/30307)[)](/langchain-ai/langchain/commit/492b4c1604300385f5fe539876f9dab7f26ab870 \"docs(readthedocs): streamline config (#30307)\") | Mar 18, 2025 |\n| [CITATION.cff](/langchain-ai/langchain/blob/master/CITATION.cff \"CITATION.cff\") | | [CITATION.cff](/langchain-ai/langchain/blob/master/CITATION.cff \"CITATION.cff\") | [rename repo namespace to langchain-ai (](/langchain-ai/langchain/commit/a363ab52923582dedc7809c6ac960201c4abe882 \"rename repo namespace to langchain-ai (#11259) ### Description renamed several repository links from `hwchase17` to `langchain-ai`. ### Why I discovered that the README file in the devcontainer contains an old repository name, so I took the opportunity to rename the old repository name in all files within the repository, excluding those that do not require changes. ### Dependencies none ### Tag maintainer @baskaryan ### Twitter handle [kzk_maeda](https://twitter.com/kzk_maeda)\")[#11259](https://github.com/langchain-ai/langchain/pull/11259)[)](/langchain-ai/langchain/commit/a363ab52923582dedc7809c6ac960201c4abe882 \"rename repo namespace to langchain-ai (#11259) ### Description renamed several repository links from `hwchase17` to `langchain-ai`. ### Why I discovered that the README file in the devcontainer contains an old repository name, so I took the opportunity to rename the old repository name in all files within the repository, excluding those that do not require changes. ### Dependencies none ### Tag maintainer @baskaryan ### Twitter handle [kzk_maeda](https://twitter.com/kzk_maeda)\") | Oct 1, 2023 |\n| [LICENSE](/langchain-ai/langchain/blob/master/LICENSE \"LICENSE\") | | [LICENSE](/langchain-ai/langchain/blob/master/LICENSE \"LICENSE\") | [Library Licenses (](/langchain-ai/langchain/commit/5eca1bd93ff42435f5d2299dd620e82d2b0c4515 \"Library Licenses (#13300) Same change as #8403 but in other libs also updates (c) LangChain Inc. instead of @hwchase17\")[#13300](https://github.com/langchain-ai/langchain/pull/13300)[)](/langchain-ai/langchain/commit/5eca1bd93ff42435f5d2299dd620e82d2b0c4515 \"Library Licenses (#13300) Same change as #8403 but in other libs also updates (c) LangChain Inc. instead of @hwchase17\") | Nov 28, 2023 |\n| [MIGRATE.md](/langchain-ai/langchain/blob/master/MIGRATE.md \"MIGRATE.md\") | | [MIGRATE.md](/langchain-ai/langchain/blob/master/MIGRATE.md \"MIGRATE.md\") | [Proofreading and Editing Report for Migration Guide (](/langchain-ai/langchain/commit/beef4c4d622d7cb79bcc532e633bffcee206b91c \"Proofreading and Editing Report for Migration Guide (#28084) Corrections and Suggestions for Migrating LangChain Code Documentation Thank you for contributing to LangChain! - [ ] **PR title**: \\\"package: description\\\" - Where \\\"package\\\" is whichever of langchain, community, core, etc. is being modified. Use \\\"docs: ...\\\" for purely docs changes, \\\"infra: ...\\\" for CI changes. - Example: \\\"community: add foobar LLM\\\" - [ ] **PR message**: ***Delete this entire checklist*** and replace with - **Description:** a description of the change - **Issue:** the issue # it fixes, if applicable - **Dependencies:** any dependencies required for this change - **Twitter handle:** if your PR gets announced, and you'd like a mention, we'll gladly shout you out! - [ ] **Add tests and docs**: If you're adding a new integration, please include 1. a test for the integration, preferably unit tests that do not rely on network access, 2. an example notebook showing its use. It lives in `docs/docs/integrations` directory. - [ ] **Lint and test**: Run `make format`, `make lint` and `make test` from the root of the package(s) you've modified. See contribution guidelines for more: https://python.langchain.com/docs/contributing/ Additional guidelines: - Make sure optional dependencies are imported within a function. - Please do not add dependencies to pyproject.toml files (even optional ones) unless they are required for unit tests. - Most PRs should not touch more than one package. - Changes should be backwards compatible. - If you are adding something to community, do not re-import it in langchain. If no one reviews your PR within a few days, please @-mention one of baskaryan, efriis, eyurtsev, ccurme, vbarda, hwchase17.\")[#28084](https://github.com/langchain-ai/langchain/pull/28084)[)](/langchain-ai/langchain/commit/beef4c4d622d7cb79bcc532e633bffcee206b91c \"Proofreading and Editing Report for Migration Guide (#28084) Corrections and Suggestions for Migrating LangChain Code Documentation Thank you for contributing to LangChain! - [ ] **PR title**: \\\"package: description\\\" - Where \\\"package\\\" is whichever of langchain, community, core, etc. is being modified. Use \\\"docs: ...\\\" for purely docs changes, \\\"infra: ...\\\" for CI changes. - Example: \\\"community: add foobar LLM\\\" - [ ] **PR message**: ***Delete this entire checklist*** and replace with - **Description:** a description of the change - **Issue:** the issue # it fixes, if applicable - **Dependencies:** any dependencies required for this change - **Twitter handle:** if your PR gets announced, and you'd like a mention, we'll gladly shout you out! - [ ] **Add tests and docs**: If you're adding a new integration, please include 1. a test for the integration, preferably unit tests that do not rely on network access, 2. an example notebook showing its use. It lives in `docs/docs/integrations` directory. - [ ] **Lint and test**: Run `make format`, `make lint` and `make test` from the root of the package(s) you've modified. See contribution guidelines for more: https://python.langchain.com/docs/contributing/ Additional guidelines: - Make sure optional dependencies are imported within a function. - Please do not add dependencies to pyproject.toml files (even optional ones) unless they are required for unit tests. - Most PRs should not touch more than one package. - Changes should be backwards compatible. - If you are adding something to community, do not re-import it in langchain. If no one reviews your PR within a few days, please @-mention one of baskaryan, efriis, eyurtsev, ccurme, vbarda, hwchase17.\") | Nov 13, 2024 |\n| [Makefile](/langchain-ai/langchain/blob/master/Makefile \"Makefile\") | | [Makefile](/langchain-ai/langchain/blob/master/Makefile \"Makefile\") | [infra: Suppress error in make api\\_docs\\_clean if index.md is missing (](/langchain-ai/langchain/commit/92af7b093355b3b326217333daa1842946c7fde7 \"infra: Suppress error in make api_docs_clean if index.md is missing (#31129) - **Description:** Added -f flag to rm when calling ```make api_docs_clean``` to suppress error by make if index.md doesn't exist. - **Dependencies:** none On calling ```make api_docs_clean``` Behavior without this PR: ``` find ./docs/api_reference -name '*_api_reference.rst' -delete git clean -fdX ./docs/api_reference rm docs/api_reference/index.md rm: cannot remove 'docs/api_reference/index.md': No such file or directory make: *** [Makefile:51: api_docs_clean] Error 1 ``` After this PR: ``` find ./docs/api_reference -name '*_api_reference.rst' -delete git clean -fdX ./docs/api_reference rm -f docs/api_reference/index.md ```\")[#β¦](https://github.com/langchain-ai/langchain/pull/31129) | May 11, 2025 |\n| [README.md](/langchain-ai/langchain/blob/master/README.md \"README.md\") | | [README.md](/langchain-ai/langchain/blob/master/README.md \"README.md\") | [docs: fix Langgraph Platform URL in Readme file (](/langchain-ai/langchain/commit/de3d8b4b5c49f79c14af0f7328f78f0d277dc48d \"docs: fix Langgraph Platform URL in Readme file (#31341) Thank you for contributing to LangChain! - [x] **PR title**: \\\"package: description\\\" - Where \\\"package\\\" is whichever of langchain, core, etc. is being modified. Use \\\"docs: ...\\\" for purely docs changes, \\\"infra: ...\\\" for CI changes. - Example: \\\"core: add foobar LLM\\\" - [x] **PR message**: ***Delete this entire checklist*** and replace with - **Description:** a description of the change - **Issue:** the issue # it fixes, if applicable - **Dependencies:** any dependencies required for this change - **Twitter handle:** if your PR gets announced, and you'd like a mention, we'll gladly shout you out! - [x] **Add tests and docs**: If you're adding a new integration, please include 1. a test for the integration, preferably unit tests that do not rely on network access, 2. an example notebook showing its use. It lives in `docs/docs/integrations` directory. - [x] **Lint and test**: Run `make format`, `make lint` and `make test` from the root of the package(s) you've modified. See contribution guidelines for more: https://python.langchain.com/docs/contributing/ Additional guidelines: - Make sure optional dependencies are imported within a function. - Please do not add dependencies to pyproject.toml files (even optional ones) unless they are required for unit tests. - Most PRs should not touch more than one package. - Changes should be backwards compatible. If no one reviews your PR within a few days, please @-mention one of baskaryan, eyurtsev, ccurme, vbarda, hwchase17.\")[#31341](https://github.com/langchain-ai/langchain/pull/31341)[)](/langchain-ai/langchain/commit/de3d8b4b5c49f79c14af0f7328f78f0d277dc48d \"docs: fix Langgraph Platform URL in Readme file (#31341) Thank you for contributing to LangChain! - [x] **PR title**: \\\"package: description\\\" - Where \\\"package\\\" is whichever of langchain, core, etc. is being modified. Use \\\"docs: ...\\\" for purely docs changes, \\\"infra: ...\\\" for CI changes. - Example: \\\"core: add foobar LLM\\\" - [x] **PR message**: ***Delete this entire checklist*** and replace with - **Description:** a description of the change - **Issue:** the issue # it fixes, if applicable - **Dependencies:** any dependencies required for this change - **Twitter handle:** if your PR gets announced, and you'd like a mention, we'll gladly shout you out! - [x] **Add tests and docs**: If you're adding a new integration, please include 1. a test for the integration, preferably unit tests that do not rely on network access, 2. an example notebook showing its use. It lives in `docs/docs/integrations` directory. - [x] **Lint and test**: Run `make format`, `make lint` and `make test` from the root of the package(s) you've modified. See contribution guidelines for more: https://python.langchain.com/docs/contributing/ Additional guidelines: - Make sure optional dependencies are imported within a function. - Please do not add dependencies to pyproject.toml files (even optional ones) unless they are required for unit tests. - Most PRs should not touch more than one package. - Changes should be backwards compatible. If no one reviews your PR within a few days, please @-mention one of baskaryan, eyurtsev, ccurme, vbarda, hwchase17.\") | May 26, 2025 |\n| [SECURITY.md](/langchain-ai/langchain/blob/master/SECURITY.md \"SECURITY.md\") | | [SECURITY.md](/langchain-ai/langchain/blob/master/SECURITY.md \"SECURITY.md\") | [fix: typo in SECURITY.md (practicies -> practices) (](/langchain-ai/langchain/commit/c25b832f514027cd9613a64e29269c2af03849e2 \"fix: typo in SECURITY.md (practicies -> practices) (#31509) **Description:** Fixes a typo in SECURITY.md (\\\"practicies\\\" β \\\"practices\\\"). Note: This PR also unifies apostrophe usage (β β '). **Issue:** N/A **Dependencies:** None **Twitter handle:** N/A Co-authored-by: δΈι εζ \")[#31509](https://github.com/langchain-ai/langchain/pull/31509)[)](/langchain-ai/langchain/commit/c25b832f514027cd9613a64e29269c2af03849e2 \"fix: typo in SECURITY.md (practicies -> practices) (#31509) **Description:** Fixes a typo in SECURITY.md (\\\"practicies\\\" β \\\"practices\\\"). Note: This PR also unifies apostrophe usage (β β '). **Issue:** N/A **Dependencies:** None **Twitter handle:** N/A Co-authored-by: δΈι εζ \") | Jun 6, 2025 |\n| [poetry.toml](/langchain-ai/langchain/blob/master/poetry.toml \"poetry.toml\") | | [poetry.toml](/langchain-ai/langchain/blob/master/poetry.toml \"poetry.toml\") | [multiple: use modern installer in poetry (](/langchain-ai/langchain/commit/007c5a85d5264203ec0e0a6c44eb13b079bfb984 \"multiple: use modern installer in poetry (#23998)\")[#23998](https://github.com/langchain-ai/langchain/pull/23998)[)](/langchain-ai/langchain/commit/007c5a85d5264203ec0e0a6c44eb13b079bfb984 \"multiple: use modern installer in poetry (#23998)\") | Jul 8, 2024 |\n| [pyproject.toml](/langchain-ai/langchain/blob/master/pyproject.toml \"pyproject.toml\") | | [pyproject.toml](/langchain-ai/langchain/blob/master/pyproject.toml \"pyproject.toml\") | [community: move to separate repo (](/langchain-ai/langchain/commit/9ff5b5d2824794c82f3df9032f09a31fe007eb19 \"community: move to separate repo (#31060) langchain-community is moving to https://github.com/langchain-ai/langchain-community\")[#31060](https://github.com/langchain-ai/langchain/pull/31060)[)](/langchain-ai/langchain/commit/9ff5b5d2824794c82f3df9032f09a31fe007eb19 \"community: move to separate repo (#31060) langchain-community is moving to https://github.com/langchain-ai/langchain-community\") | Apr 29, 2025 |\n| [uv.lock](/langchain-ai/langchain/blob/master/uv.lock \"uv.lock\") | | [uv.lock](/langchain-ai/langchain/blob/master/uv.lock \"uv.lock\") | [core: Add ruff rules RUF (](/langchain-ai/langchain/commit/a8f2ddee318ca29434be8be0dfba396a0d9ad00b \"core: Add ruff rules RUF (#29353) See https://docs.astral.sh/ruff/rules/#ruff-specific-rules-ruf Mostly: * [RUF022](https://docs.astral.sh/ruff/rules/unsorted-dunder-all/) (unsorted `__all__`) * [RUF100](https://docs.astral.sh/ruff/rules/unused-noqa/) (unused noqa) * [RUF021](https://docs.astral.sh/ruff/rules/parenthesize-chained-operators/) (parenthesize-chained-operators) * [RUF015](https://docs.astral.sh/ruff/rules/unnecessary-iterable-allocation-for-first-element/) (unnecessary-iterable-allocation-for-first-element) * [RUF005](https://docs.astral.sh/ruff/rules/collection-literal-concatenation/) (collection-literal-concatenation) * [RUF046](https://docs.astral.sh/ruff/rules/unnecessary-cast-to-int/) (unnecessary-cast-to-int) --------- Co-authored-by: Eugene Yurtsev \")[#29353](https://github.com/langchain-ai/langchain/pull/29353)[)](/langchain-ai/langchain/commit/a8f2ddee318ca29434be8be0dfba396a0d9ad00b \"core: Add ruff rules RUF (#29353) See https://docs.astral.sh/ruff/rules/#ruff-specific-rules-ruf Mostly: * [RUF022](https://docs.astral.sh/ruff/rules/unsorted-dunder-all/) (unsorted `__all__`) * [RUF100](https://docs.astral.sh/ruff/rules/unused-noqa/) (unused noqa) * [RUF021](https://docs.astral.sh/ruff/rules/parenthesize-chained-operators/) (parenthesize-chained-operators) * [RUF015](https://docs.astral.sh/ruff/rules/unnecessary-iterable-allocation-for-first-element/) (unnecessary-iterable-allocation-for-first-element) * [RUF005](https://docs.astral.sh/ruff/rules/collection-literal-concatenation/) (collection-literal-concatenation) * [RUF046](https://docs.astral.sh/ruff/rules/unnecessary-cast-to-int/) (unnecessary-cast-to-int) --------- Co-authored-by: Eugene Yurtsev \") | May 15, 2025 |\n| [yarn.lock](/langchain-ai/langchain/blob/master/yarn.lock \"yarn.lock\") | | [yarn.lock](/langchain-ai/langchain/blob/master/yarn.lock \"yarn.lock\") | [box: add langchain box package and DocumentLoader (](/langchain-ai/langchain/commit/55fd2e21584d6ef269ff3bdd2f2b7bd49889ee3c \"box: add langchain box package and DocumentLoader (#25506) Thank you for contributing to LangChain! -Description: Adding new package: `langchain-box`: * `langchain_box.document_loaders.BoxLoader` β DocumentLoader functionality * `langchain_box.utilities.BoxAPIWrapper` β Box-specific code * `langchain_box.utilities.BoxAuth` β Helper class for Box authentication * `langchain_box.utilities.BoxAuthType` β enum used by BoxAuth class - Twitter handle: @boxplatform - [x] **Add tests and docs**: If you're adding a new integration, please include 1. a test for the integration, preferably unit tests that do not rely on network access, 2. an example notebook showing its use. It lives in `docs/docs/integrations` directory. - [x] **Lint and test**: Run `make format`, `make lint` and `make test` from the root of the package(s) you've modified. See contribution guidelines for more: https://python.langchain.com/docs/contributing/ Additional guidelines: - Make sure optional dependencies are imported within a function. - Please do not add dependencies to pyproject.toml files (even optional ones) unless they are required for unit tests. - Most PRs should not touch more than one package. - Changes should be backwards compatible. - If you are adding something to community, do not re-import it in langchain. If no one reviews your PR within a few days, please @-mention one of baskaryan, efriis, eyurtsev, ccurme, vbarda, hwchase17. --------- Co-authored-by: Erick Friis Co-authored-by: Erick Friis \")[#25506](https://github.com/langchain-ai/langchain/pull/25506)[)](/langchain-ai/langchain/commit/55fd2e21584d6ef269ff3bdd2f2b7bd49889ee3c \"box: add langchain box package and DocumentLoader (#25506) Thank you for contributing to LangChain! -Description: Adding new package: `langchain-box`: * `langchain_box.document_loaders.BoxLoader` β DocumentLoader functionality * `langchain_box.utilities.BoxAPIWrapper` β Box-specific code * `langchain_box.utilities.BoxAuth` β Helper class for Box authentication * `langchain_box.utilities.BoxAuthType` β enum used by BoxAuth class - Twitter handle: @boxplatform - [x] **Add tests and docs**: If you're adding a new integration, please include 1. a test for the integration, preferably unit tests that do not rely on network access, 2. an example notebook showing its use. It lives in `docs/docs/integrations` directory. - [x] **Lint and test**: Run `make format`, `make lint` and `make test` from the root of the package(s) you've modified. See contribution guidelines for more: https://python.langchain.com/docs/contributing/ Additional guidelines: - Make sure optional dependencies are imported within a function. - Please do not add dependencies to pyproject.toml files (even optional ones) unless they are required for unit tests. - Most PRs should not touch more than one package. - Changes should be backwards compatible. - If you are adding something to community, do not re-import it in langchain. If no one reviews your PR within a few days, please @-mention one of baskaryan, efriis, eyurtsev, ccurme, vbarda, hwchase17. --------- Co-authored-by: Erick Friis Co-authored-by: Erick Friis \") | Aug 20, 2024 |\n| View all files | | |\n\n## Repository files navigation\n\n- [README](#)\n- [Code of conduct](#)\n- [MIT license](#)\n- [Security](#)\n\n\n\n[](https://github.com/langchain-ai/langchain/releases)\n[](https://github.com/langchain-ai/langchain/actions/workflows/check_diffs.yml)\n[](https://opensource.org/licenses/MIT)\n[](https://pypistats.org/packages/langchain-core)\n[](https://star-history.com/#langchain-ai/langchain)\n[](https://github.com/langchain-ai/langchain/issues)\n[](https://vscode.dev/redirect?url=vscode://ms-vscode-remote.remote-containers/cloneInVolume?url=https://github.com/langchain-ai/langchain)\n[![](https://github.com/codespaces/badge.svg \"Open in Github Codespace\")](https://codespaces.new/langchain-ai/langchain)\n[](https://twitter.com/langchainai)\n[](https://codspeed.io/langchain-ai/langchain)\n\nNote\n\nLooking for the JS/TS library? Check out [LangChain.js](https://github.com/langchain-ai/langchainjs).\n\nLangChain is a framework for building LLM-powered applications. It helps you chain\ntogether interoperable components and third-party integrations to simplify AI\napplication development β all while future-proofing decisions as the underlying\ntechnology evolves.\n\n```\npip install -U langchain\n```\n\nTo learn more about LangChain, check out\n[the docs](https://python.langchain.com/docs/introduction/). If youβre looking for more\nadvanced customization or agent orchestration, check out\n[LangGraph](https://langchain-ai.github.io/langgraph/), our framework for building\ncontrollable agent workflows.\n\n## Why use LangChain?\n\nLangChain helps developers build applications powered by LLMs through a standard\ninterface for models, embeddings, vector stores, and more.\n\nUse LangChain for:\n\n- **Real-time data augmentation**. Easily connect LLMs to diverse data sources and\n external / internal systems, drawing from LangChainβs vast library of integrations with\n model providers, tools, vector stores, retrievers, and more.\n- **Model interoperability**. Swap models in and out as your engineering team\n experiments to find the best choice for your applicationβs needs. As the industry\n frontier evolves, adapt quickly β LangChainβs abstractions keep you moving without\n losing momentum.\n\n## LangChainβs ecosystem\n\nWhile the LangChain framework can be used standalone, it also integrates seamlessly\nwith any LangChain product, giving developers a full suite of tools when building LLM\napplications.\n\nTo improve your LLM application development, pair LangChain with:\n\n- [LangSmith](http://www.langchain.com/langsmith) - Helpful for agent evals and\n observability. Debug poor-performing LLM app runs, evaluate agent trajectories, gain\n visibility in production, and improve performance over time.\n- [LangGraph](https://langchain-ai.github.io/langgraph/) - Build agents that can\n reliably handle complex tasks with LangGraph, our low-level agent orchestration\n framework. LangGraph offers customizable architecture, long-term memory, and\n human-in-the-loop workflows β and is trusted in production by companies like LinkedIn,\n Uber, Klarna, and GitLab.\n- [LangGraph Platform](https://langchain-ai.github.io/langgraph/concepts/langgraph_platform/) - Deploy\n and scale agents effortlessly with a purpose-built deployment platform for long\n running, stateful workflows. Discover, reuse, configure, and share agents across\n teams β and iterate quickly with visual prototyping in\n [LangGraph Studio](https://langchain-ai.github.io/langgraph/concepts/langgraph_studio/).\n\n## Additional resources\n\n- [Tutorials](https://python.langchain.com/docs/tutorials/): Simple walkthroughs with\n guided examples on getting started with LangChain.\n- [How-to Guides](https://python.langchain.com/docs/how_to/): Quick, actionable code\n snippets for topics such as tool calling, RAG use cases, and more.\n- [Conceptual Guides](https://python.langchain.com/docs/concepts/): Explanations of key\n concepts behind the LangChain framework.\n- [API Reference](https://python.langchain.com/api_reference/): Detailed reference on\n navigating base packages and integrations for LangChain.\n\n## About\n\nπ¦π Build context-aware reasoning applications\n\n[python.langchain.com](https://python.langchain.com \"https://python.langchain.com\")\n\n### Resources\n\n[Readme](#readme-ov-file)\n\n### License\n\n[MIT license](#MIT-1-ov-file)\n\n### Code of conduct\n\n[Code of conduct](#coc-ov-file)\n\n### Security policy\n\n[Security policy](#security-ov-file)\n\n### Citation\n\nCite this repository\n\n\n\nLoading\n\nSomething went wrong.\n\n### Uh oh!\n\nThere was an error while loading. Please reload this page.\n\n[Activity](/langchain-ai/langchain/activity)\n\n[Custom properties](/langchain-ai/langchain/custom-properties)\n\n### Stars\n\n[**109k**\nstars](/langchain-ai/langchain/stargazers)\n\n### Watchers\n\n[**759**\nwatching](/langchain-ai/langchain/watchers)\n\n### Forks\n\n[**17.8k**\nforks](/langchain-ai/langchain/forks)\n\n[Report repository](/contact/report-content?content_url=https%3A%2F%2Fgithub.com%2Flangchain-ai%2Flangchain&report=langchain-ai+%28user%29)\n\n## [Releases 905](/langchain-ai/langchain/releases)\n\n[langchain-core==0.3.65\n\nLatest\n\nJun 10, 2025](/langchain-ai/langchain/releases/tag/langchain-core%3D%3D0.3.65)\n\n[+ 904 releases](/langchain-ai/langchain/releases)\n\n## [Packages 0](/orgs/langchain-ai/packages?repo_name=langchain)\n\nNo packages published\n\n## [Contributors 3,640](/langchain-ai/langchain/graphs/contributors)\n\n- [](https://github.com/baskaryan)\n- [](https://github.com/hwchase17)\n- [](https://github.com/ccurme)\n- [](https://github.com/eyurtsev)\n- [](https://github.com/nfcampos)\n- [](https://github.com/leo-gan)\n- [](https://github.com/hinthornw)\n- [](https://github.com/dev2049)\n- [](https://github.com/cbornet)\n- [](https://github.com/rlancemartin)\n- [](https://github.com/vowelparrot)\n- [](https://github.com/jacoblee93)\n- [](https://github.com/tomasonjo)\n- [](https://github.com/obi1kenobi)\n\n[+ 3,626 contributors](/langchain-ai/langchain/graphs/contributors)\n\n## Languages\n\n- [Jupyter Notebook\n 79.5%](/langchain-ai/langchain/search?l=jupyter-notebook)\n- [Python\n 20.3%](/langchain-ai/langchain/search?l=python)\n- [MDX\n 0.1%](/langchain-ai/langchain/search?l=mdx)\n- [Makefile\n 0.1%](/langchain-ai/langchain/search?l=makefile)\n- [Shell\n 0.0%](/langchain-ai/langchain/search?l=shell)\n- [XSLT\n 0.0%](/langchain-ai/langchain/search?l=xslt)\n\n## Footer\n\nΒ© 2025 GitHub,Β Inc.\n\n### Footer navigation\n\n- [Terms](https://docs.github.com/site-policy/github-terms/github-terms-of-service)\n- [Privacy](https://docs.github.com/site-policy/privacy-policies/github-privacy-statement)\n- [Security](https://github.com/security)\n- [Status](https://www.githubstatus.com/)\n- [Docs](https://docs.github.com/)\n- [Contact](https://support.github.com?tags=dotcom-footer)\n- Manage cookies\n- Do not share my personal information\n\n\n\nYou canβt perform that action at this time."}, {"url": "https://github.com/langchain-ai/langgraph", "title": "langchain-ai/langgraph: Build resilient language agents as graphs.", "content": "GitHub - langchain-ai/langgraph: Build resilient language agents as graphs. langchain-ai.github.io/langgraph/ Trusted by companies shaping the future of agents β including Klarna, Replit, Elastic, and more β LangGraph is a low-level orchestration framework for building, managing, and deploying long-running, stateful agents. Or, to learn how to build an agent workflow with a customizable architecture, long-term memory, and other complex task handling, see the LangGraph basics tutorials. LangGraph provides low-level supporting infrastructure for _any_ long-running, stateful workflow or agent. While LangGraph can be used standalone, it also integrates seamlessly with any LangChain product, giving developers a full suite of tools for building agents. * LangGraph Platform β Deploy and scale agents effortlessly with a purpose-built deployment platform for long running, stateful workflows. langchain-ai.github.io/langgraph/", "score": 0.03588415, "raw_content": "GitHub - langchain-ai/langgraph: Build resilient language agents as graphs.\n\n===============\n\n[Skip to content](https://github.com/langchain-ai/langgraph#start-of-content)\nNavigation Menu\n---------------\n\nToggle navigation\n\n[](https://github.com/)\n\n[Sign in](https://github.com/login?return_to=https%3A%2F%2Fgithub.com%2Flangchain-ai%2Flanggraph)\n\nAppearance settings\n\n* Product \n\n * [GitHub Copilot Write better code with AI](https://github.com/features/copilot)\n * [GitHub Models New Manage and compare prompts](https://github.com/features/models)\n * [GitHub Advanced Security Find and fix vulnerabilities](https://github.com/security/advanced-security)\n * [Actions Automate any workflow](https://github.com/features/actions)\n * [Codespaces Instant dev environments](https://github.com/features/codespaces)\n\n * [Issues Plan and track work](https://github.com/features/issues)\n * [Code Review Manage code changes](https://github.com/features/code-review)\n * [Discussions Collaborate outside of code](https://github.com/features/discussions)\n * [Code Search Find more, search less](https://github.com/features/code-search)\n\nExplore\n * [Why GitHub](https://github.com/why-github)\n * [All features](https://github.com/features)\n * [Documentation](https://docs.github.com/)\n * [GitHub Skills](https://skills.github.com/)\n * [Blog](https://github.blog/)\n\n* Solutions \n\nBy company size\n * [Enterprises](https://github.com/enterprise)\n * [Small and medium teams](https://github.com/team)\n * [Startups](https://github.com/enterprise/startups)\n * [Nonprofits](https://github.com/solutions/industry/nonprofits)\n\nBy use case\n * [DevSecOps](https://github.com/solutions/use-case/devsecops)\n * [DevOps](https://github.com/solutions/use-case/devops)\n * [CI/CD](https://github.com/solutions/use-case/ci-cd)\n * [View all use cases](https://github.com/solutions/use-case)\n\nBy industry\n * [Healthcare](https://github.com/solutions/industry/healthcare)\n * [Financial services](https://github.com/solutions/industry/financial-services)\n * [Manufacturing](https://github.com/solutions/industry/manufacturing)\n * [Government](https://github.com/solutions/industry/government)\n * [View all industries](https://github.com/solutions/industry)\n\n[View all solutions](https://github.com/solutions)\n\n* Resources \n\nTopics\n * [AI](https://github.com/resources/articles/ai)\n * [DevOps](https://github.com/resources/articles/devops)\n * [Security](https://github.com/resources/articles/security)\n * [Software Development](https://github.com/resources/articles/software-development)\n * [View all](https://github.com/resources/articles)\n\nExplore\n * [Learning Pathways](https://resources.github.com/learn/pathways)\n * [Events & Webinars](https://resources.github.com/)\n * [Ebooks & Whitepapers](https://github.com/resources/whitepapers)\n * [Customer Stories](https://github.com/customer-stories)\n * [Partners](https://partner.github.com/)\n * [Executive Insights](https://github.com/solutions/executive-insights)\n\n* Open Source \n\n * [GitHub Sponsors Fund open source developers](https://github.com/sponsors)\n\n * [The ReadME Project GitHub community articles](https://github.com/readme)\n\nRepositories\n * [Topics](https://github.com/topics)\n * [Trending](https://github.com/trending)\n * [Collections](https://github.com/collections)\n\n* Enterprise \n\n * [Enterprise platform AI-powered developer platform](https://github.com/enterprise)\n\nAvailable add-ons\n * [GitHub Advanced Security Enterprise-grade security features](https://github.com/security/advanced-security)\n * [Copilot for business Enterprise-grade AI features](https://github.com/features/copilot/copilot-business)\n * [Premium Support Enterprise-grade 24/7 support](https://github.com/premium-support)\n\n* [Pricing](https://github.com/pricing)\n\nSearch or jump to...\n\nSearch code, repositories, users, issues, pull requests...\n==========================================================\n\n Search \n\nClear\n\n[Search syntax tips](https://docs.github.com/search-github/github-code-search/understanding-github-code-search-syntax)\n\nProvide feedback\n================\n\nWe read every piece of feedback, and take your input very seriously.\n\n- [x] Include my email address so I can be contacted \n\n Cancel Submit feedback \n\nSaved searches\n==============\n\nUse saved searches to filter your results more quickly\n------------------------------------------------------\n\nName \n\nQuery \n\nTo see all available qualifiers, see our [documentation](https://docs.github.com/search-github/github-code-search/understanding-github-code-search-syntax).\n\n Cancel Create saved search \n\n[Sign in](https://github.com/login?return_to=https%3A%2F%2Fgithub.com%2Flangchain-ai%2Flanggraph)\n\n[Sign up](https://github.com/signup?ref_cta=Sign+up&ref_loc=header+logged+out&ref_page=%2F%3Cuser-name%3E%2F%3Crepo-name%3E&source=header-repo&source_repo=langchain-ai%2Flanggraph)\n\nAppearance settings\n\nResetting focus\n\nYou signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert\n\n{{ message }}\n\n[langchain-ai](https://github.com/langchain-ai)/**[langgraph](https://github.com/langchain-ai/langgraph)**Public\n\n* [Notifications](https://github.com/login?return_to=%2Flangchain-ai%2Flanggraph)You must be signed in to change notification settings\n* [Fork 2.4k](https://github.com/login?return_to=%2Flangchain-ai%2Flanggraph)\n* [Star 14.5k](https://github.com/login?return_to=%2Flangchain-ai%2Flanggraph) \n\nBuild resilient language agents as graphs.\n\n[langchain-ai.github.io/langgraph/](https://langchain-ai.github.io/langgraph/ \"https://langchain-ai.github.io/langgraph/\")\n\n### License\n\n[MIT license](https://github.com/langchain-ai/langgraph/blob/main/LICENSE)\n\n[14.5k stars](https://github.com/langchain-ai/langgraph/stargazers)[2.4k forks](https://github.com/langchain-ai/langgraph/forks)[Branches](https://github.com/langchain-ai/langgraph/branches)[Tags](https://github.com/langchain-ai/langgraph/tags)[Activity](https://github.com/langchain-ai/langgraph/activity)\n\n[Star](https://github.com/login?return_to=%2Flangchain-ai%2Flanggraph)\n\n[Notifications](https://github.com/login?return_to=%2Flangchain-ai%2Flanggraph)You must be signed in to change notification settings\n\n* [Code](https://github.com/langchain-ai/langgraph)\n* [Issues 75](https://github.com/langchain-ai/langgraph/issues)\n* [Pull requests 26](https://github.com/langchain-ai/langgraph/pulls)\n* [Discussions](https://github.com/langchain-ai/langgraph/discussions)\n* [Actions](https://github.com/langchain-ai/langgraph/actions)\n* [Projects 2](https://github.com/langchain-ai/langgraph/projects)\n* [Security](https://github.com/langchain-ai/langgraph/security)[](https://github.com/langchain-ai/langgraph/security)[](https://github.com/langchain-ai/langgraph/security)[](https://github.com/langchain-ai/langgraph/security)[### Uh oh!](https://github.com/langchain-ai/langgraph/security)\n[There was an error while loading.](https://github.com/langchain-ai/langgraph/security)Please reload this page. \n* [Insights](https://github.com/langchain-ai/langgraph/pulse)\n\nAdditional navigation options\n\n* [Code](https://github.com/langchain-ai/langgraph)\n* [Issues](https://github.com/langchain-ai/langgraph/issues)\n* [Pull requests](https://github.com/langchain-ai/langgraph/pulls)\n* [Discussions](https://github.com/langchain-ai/langgraph/discussions)\n* [Actions](https://github.com/langchain-ai/langgraph/actions)\n* [Projects](https://github.com/langchain-ai/langgraph/projects)\n* [Security](https://github.com/langchain-ai/langgraph/security)\n* [Insights](https://github.com/langchain-ai/langgraph/pulse)\n\nlangchain-ai/langgraph\n======================\n\nmain\n\n[**102**Branches](https://github.com/langchain-ai/langgraph/branches)[**353**Tags](https://github.com/langchain-ai/langgraph/tags)\n\n[](https://github.com/langchain-ai/langgraph/branches)[](https://github.com/langchain-ai/langgraph/tags)\n\nGo to file\n\nCode\n\nOpen more actions menu\n\nFolders and files\n-----------------\n\n| Name | Name | Last commit message | Last commit date |\n| --- | --- | --- | --- |\n| Latest commit ------------- [](https://github.com/andrewnguonly)[andrewnguonly](https://github.com/langchain-ai/langgraph/commits?author=andrewnguonly) [docs: Remove note for](https://github.com/langchain-ai/langgraph/commit/4a252bd03a67c3ca857f7a7e4ffc0d589057415c)`LANGSMITH_TRACING`[environment variable (](https://github.com/langchain-ai/langgraph/commit/4a252bd03a67c3ca857f7a7e4ffc0d589057415c)[#5147](https://github.com/langchain-ai/langgraph/pull/5147)[)](https://github.com/langchain-ai/langgraph/commit/4a252bd03a67c3ca857f7a7e4ffc0d589057415c) Open commit details success Jun 20, 2025 [4a252bd](https://github.com/langchain-ai/langgraph/commit/4a252bd03a67c3ca857f7a7e4ffc0d589057415c)Β·Jun 20, 2025 History ------- [5,726 Commits](https://github.com/langchain-ai/langgraph/commits/main/) Open commit details [](https://github.com/langchain-ai/langgraph/commits/main/) |\n| [.github](https://github.com/langchain-ai/langgraph/tree/main/.github \".github\") | [.github](https://github.com/langchain-ai/langgraph/tree/main/.github \".github\") | [oss: auto apply labels to contributor issues (](https://github.com/langchain-ai/langgraph/commit/45cd4e1928eb266c973bb082adcf367e2c6f92d2 \"oss: auto apply labels to contributor issues (#5067) auto apply labels\")[#5067](https://github.com/langchain-ai/langgraph/pull/5067)[)](https://github.com/langchain-ai/langgraph/commit/45cd4e1928eb266c973bb082adcf367e2c6f92d2 \"oss: auto apply labels to contributor issues (#5067) auto apply labels\") | Jun 11, 2025 |\n| [docs](https://github.com/langchain-ai/langgraph/tree/main/docs \"docs\") | [docs](https://github.com/langchain-ai/langgraph/tree/main/docs \"docs\") | [docs: Remove note for](https://github.com/langchain-ai/langgraph/commit/4a252bd03a67c3ca857f7a7e4ffc0d589057415c \"docs: Remove note for `LANGSMITH_TRACING` environment variable (#5147) Remove note for LANGSMITH_TRACING.\")`LANGSMITH_TRACING`[environment variable (](https://github.com/langchain-ai/langgraph/commit/4a252bd03a67c3ca857f7a7e4ffc0d589057415c \"docs: Remove note for `LANGSMITH_TRACING` environment variable (#5147) Remove note for LANGSMITH_TRACING.\")[#5147](https://github.com/langchain-ai/langgraph/pull/5147)[)](https://github.com/langchain-ai/langgraph/commit/4a252bd03a67c3ca857f7a7e4ffc0d589057415c \"docs: Remove note for `LANGSMITH_TRACING` environment variable (#5147) Remove note for LANGSMITH_TRACING.\") | Jun 20, 2025 |\n| [examples](https://github.com/langchain-ai/langgraph/tree/main/examples \"examples\") | [examples](https://github.com/langchain-ai/langgraph/tree/main/examples \"examples\") | [docs: Replace unsupported models with structured output-supported modβ¦](https://github.com/langchain-ai/langgraph/commit/0035ab9825cdfc588d250474e6da657adc2b39c9 \"docs: Replace unsupported models with structured output-supported models (#3982)\") | Jun 9, 2025 |\n| [libs](https://github.com/langchain-ai/langgraph/tree/main/libs \"libs\") | [libs](https://github.com/langchain-ai/langgraph/tree/main/libs \"libs\") | [Reduce extraneous keys in checkpoint.metadata](https://github.com/langchain-ai/langgraph/commit/a1c856c088712147e76eaffdd67722df82891260 \"Reduce extraneous keys in checkpoint.metadata - Leave it up to each checkpointer implementation to decide whether to merge in configurable/metadata (previously PregelLoop would do some of this always) - Never copy over internal langgraph keys into checkpoint.metadata (these are redundant/misleading to include)\") | Jun 18, 2025 |\n| [.gitignore](https://github.com/langchain-ai/langgraph/blob/main/.gitignore \".gitignore\") | [.gitignore](https://github.com/langchain-ai/langgraph/blob/main/.gitignore \".gitignore\") | [Improve type checking on graph](https://github.com/langchain-ai/langgraph/commit/37c215b5a2deb16ae042eba0380407bf1a8d0dd5 \"Improve type checking on graph `init` and `invoke`/`stream` (#4932)\")`init`[and](https://github.com/langchain-ai/langgraph/commit/37c215b5a2deb16ae042eba0380407bf1a8d0dd5 \"Improve type checking on graph `init` and `invoke`/`stream` (#4932)\")`invoke`[/](https://github.com/langchain-ai/langgraph/commit/37c215b5a2deb16ae042eba0380407bf1a8d0dd5 \"Improve type checking on graph `init` and `invoke`/`stream` (#4932)\")`stream`[(](https://github.com/langchain-ai/langgraph/commit/37c215b5a2deb16ae042eba0380407bf1a8d0dd5 \"Improve type checking on graph `init` and `invoke`/`stream` (#4932)\")[#4932](https://github.com/langchain-ai/langgraph/pull/4932)[)](https://github.com/langchain-ai/langgraph/commit/37c215b5a2deb16ae042eba0380407bf1a8d0dd5 \"Improve type checking on graph `init` and `invoke`/`stream` (#4932)\") | Jun 3, 2025 |\n| [.markdown-link-check.config.json](https://github.com/langchain-ai/langgraph/blob/main/.markdown-link-check.config.json \".markdown-link-check.config.json\") | [.markdown-link-check.config.json](https://github.com/langchain-ai/langgraph/blob/main/.markdown-link-check.config.json \".markdown-link-check.config.json\") | [Use create_react_agent (](https://github.com/langchain-ai/langgraph/commit/2c32a38c4206ae029df1ba308617846ad52ba0bb \"Use create_react_agent (#441)\")[#441](https://github.com/langchain-ai/langgraph/pull/441)[)](https://github.com/langchain-ai/langgraph/commit/2c32a38c4206ae029df1ba308617846ad52ba0bb \"Use create_react_agent (#441)\") | May 13, 2024 |\n| [AGENTS.md](https://github.com/langchain-ai/langgraph/blob/main/AGENTS.md \"AGENTS.md\") | [AGENTS.md](https://github.com/langchain-ai/langgraph/blob/main/AGENTS.md \"AGENTS.md\") | [Refine dependency map](https://github.com/langchain-ai/langgraph/commit/b95267a3cc80f612e4e0591f2b8d8134d3a89cb4 \"Refine dependency map\") | Jun 10, 2025 |\n| [CONTRIBUTING.md](https://github.com/langchain-ai/langgraph/blob/main/CONTRIBUTING.md \"CONTRIBUTING.md\") | [CONTRIBUTING.md](https://github.com/langchain-ai/langgraph/blob/main/CONTRIBUTING.md \"CONTRIBUTING.md\") | [docs: fix a typo (](https://github.com/langchain-ai/langgraph/commit/f8b5e05c47cb553fd4ff65a9693a0805c3505ca6 \"docs: fix a typo (#4908)\")[#4908](https://github.com/langchain-ai/langgraph/pull/4908)[)](https://github.com/langchain-ai/langgraph/commit/f8b5e05c47cb553fd4ff65a9693a0805c3505ca6 \"docs: fix a typo (#4908)\") | Jun 2, 2025 |\n| [LICENSE](https://github.com/langchain-ai/langgraph/blob/main/LICENSE \"LICENSE\") | [LICENSE](https://github.com/langchain-ai/langgraph/blob/main/LICENSE \"LICENSE\") | [libs: add cli, sdk-py, sdk-js and move core langgraph](https://github.com/langchain-ai/langgraph/commit/1a06d500d4282cfdb2ae9d7748bb570e8162acdf \"libs: add cli, sdk-py, sdk-js and move core langgraph\") | Jun 18, 2024 |\n| [Makefile](https://github.com/langchain-ai/langgraph/blob/main/Makefile \"Makefile\") | [Makefile](https://github.com/langchain-ai/langgraph/blob/main/Makefile \"Makefile\") | [lint: use](https://github.com/langchain-ai/langgraph/commit/5e7566f4a37d663cf666240f60fd0e1d61a43446 \"lint: use `pep 604` union syntax and `pep 585` generic syntax (#4963) * new union syntax * fix test * second round of conversions by injecting future annotations * format + add top level makefile\")`pep 604`[union syntax and](https://github.com/langchain-ai/langgraph/commit/5e7566f4a37d663cf666240f60fd0e1d61a43446 \"lint: use `pep 604` union syntax and `pep 585` generic syntax (#4963) * new union syntax * fix test * second round of conversions by injecting future annotations * format + add top level makefile\")`pep 585`[generic syntax (](https://github.com/langchain-ai/langgraph/commit/5e7566f4a37d663cf666240f60fd0e1d61a43446 \"lint: use `pep 604` union syntax and `pep 585` generic syntax (#4963) * new union syntax * fix test * second round of conversions by injecting future annotations * format + add top level makefile\")[#4963](https://github.com/langchain-ai/langgraph/pull/4963)[)](https://github.com/langchain-ai/langgraph/commit/5e7566f4a37d663cf666240f60fd0e1d61a43446 \"lint: use `pep 604` union syntax and `pep 585` generic syntax (#4963) * new union syntax * fix test * second round of conversions by injecting future annotations * format + add top level makefile\") | Jun 5, 2025 |\n| [README.md](https://github.com/langchain-ai/langgraph/blob/main/README.md \"README.md\") | [README.md](https://github.com/langchain-ai/langgraph/blob/main/README.md \"README.md\") | [Remove gitmcp badge (](https://github.com/langchain-ai/langgraph/commit/d467ec6556498c2ce623f90ab2e9422aafeee530 \"Remove gitmcp badge (#5055) * Remove gitmcp badge * xt * x\")[#5055](https://github.com/langchain-ai/langgraph/pull/5055)[)](https://github.com/langchain-ai/langgraph/commit/d467ec6556498c2ce623f90ab2e9422aafeee530 \"Remove gitmcp badge (#5055) * Remove gitmcp badge * xt * x\") | Jun 11, 2025 |\n| [security.md](https://github.com/langchain-ai/langgraph/blob/main/security.md \"security.md\") | [security.md](https://github.com/langchain-ai/langgraph/blob/main/security.md \"security.md\") | [x (](https://github.com/langchain-ai/langgraph/commit/76b6394e7f7b3af10bc054e09715a6da1aa9968a \"x (#210)\")[#210](https://github.com/langchain-ai/langgraph/pull/210)[)](https://github.com/langchain-ai/langgraph/commit/76b6394e7f7b3af10bc054e09715a6da1aa9968a \"x (#210)\") | Mar 15, 2024 |\n| View all files |\n\nRepository files navigation\n---------------------------\n\n* [README](https://github.com/langchain-ai/langgraph#)\n* [MIT license](https://github.com/langchain-ai/langgraph#)\n* [Security](https://github.com/langchain-ai/langgraph#)\n\n\n\n[](https://pypi.org/project/langgraph/)[](https://pepy.tech/project/langgraph)[](https://github.com/langchain-ai/langgraph/issues)[](https://langchain-ai.github.io/langgraph/)\n\nTrusted by companies shaping the future of agents β including Klarna, Replit, Elastic, and more β LangGraph is a low-level orchestration framework for building, managing, and deploying long-running, stateful agents.\n\nGet started\n-----------\n\n[](https://github.com/langchain-ai/langgraph#get-started)\n\nInstall LangGraph:\n\n```\npip install -U langgraph\n```\n\nThen, create an agent [using prebuilt components](https://langchain-ai.github.io/langgraph/agents/agents/):\n\nundefinedpython\n# pip install -qU \"langchain[anthropic]\" to call the model\n\nfrom langgraph.prebuilt import create_react_agent\n\ndef get_weather(city: str) -> str:\n \"\"\"Get weather for a given city.\"\"\"\n return f\"It's always sunny in {city}!\"\n\nagent = create_react_agent(\n model=\"anthropic:claude-3-7-sonnet-latest\",\n tools=[get_weather],\n prompt=\"You are a helpful assistant\"\n)\n\n# Run the agent\nagent.invoke(\n {\"messages\": [{\"role\": \"user\", \"content\": \"what is the weather in sf\"}]}\n)\nundefined\n\nFor more information, see the [Quickstart](https://langchain-ai.github.io/langgraph/agents/agents/). Or, to learn how to build an [agent workflow](https://langchain-ai.github.io/langgraph/concepts/low_level/) with a customizable architecture, long-term memory, and other complex task handling, see the [LangGraph basics tutorials](https://langchain-ai.github.io/langgraph/tutorials/get-started/1-build-basic-chatbot/).\n\nCore benefits\n-------------\n\n[](https://github.com/langchain-ai/langgraph#core-benefits)\n\nLangGraph provides low-level supporting infrastructure for _any_ long-running, stateful workflow or agent. LangGraph does not abstract prompts or architecture, and provides the following central benefits:\n\n* [Durable execution](https://langchain-ai.github.io/langgraph/concepts/durable_execution/): Build agents that persist through failures and can run for extended periods, automatically resuming from exactly where they left off.\n* [Human-in-the-loop](https://langchain-ai.github.io/langgraph/concepts/human_in_the_loop/): Seamlessly incorporate human oversight by inspecting and modifying agent state at any point during execution.\n* [Comprehensive memory](https://langchain-ai.github.io/langgraph/concepts/memory/): Create truly stateful agents with both short-term working memory for ongoing reasoning and long-term persistent memory across sessions.\n* [Debugging with LangSmith](http://www.langchain.com/langsmith): Gain deep visibility into complex agent behavior with visualization tools that trace execution paths, capture state transitions, and provide detailed runtime metrics.\n* [Production-ready deployment](https://langchain-ai.github.io/langgraph/concepts/deployment_options/): Deploy sophisticated agent systems confidently with scalable infrastructure designed to handle the unique challenges of stateful, long-running workflows.\n\nLangGraphβs ecosystem\n---------------------\n\n[](https://github.com/langchain-ai/langgraph#langgraphs-ecosystem)\n\nWhile LangGraph can be used standalone, it also integrates seamlessly with any LangChain product, giving developers a full suite of tools for building agents. To improve your LLM application development, pair LangGraph with:\n\n* [LangSmith](http://www.langchain.com/langsmith) β Helpful for agent evals and observability. Debug poor-performing LLM app runs, evaluate agent trajectories, gain visibility in production, and improve performance over time.\n* [LangGraph Platform](https://langchain-ai.github.io/langgraph/concepts/#langgraph-platform) β Deploy and scale agents effortlessly with a purpose-built deployment platform for long running, stateful workflows. Discover, reuse, configure, and share agents across teams β and iterate quickly with visual prototyping in [LangGraph Studio](https://langchain-ai.github.io/langgraph/concepts/langgraph_studio/).\n* [LangChain](https://python.langchain.com/docs/introduction/) β Provides integrations and composable components to streamline LLM application development.\n\nNote\n\nLooking for the JS version of LangGraph? See the [JS repo](https://github.com/langchain-ai/langgraphjs) and the [JS docs](https://langchain-ai.github.io/langgraphjs/).\n\nAdditional resources\n--------------------\n\n[](https://github.com/langchain-ai/langgraph#additional-resources)\n\n* [Guides](https://langchain-ai.github.io/langgraph/how-tos/): Quick, actionable code snippets for topics such as streaming, adding memory & persistence, and design patterns (e.g. branching, subgraphs, etc.).\n* [Reference](https://langchain-ai.github.io/langgraph/reference/graphs/): Detailed reference on core classes, methods, how to use the graph and checkpointing APIs, and higher-level prebuilt components.\n* [Examples](https://langchain-ai.github.io/langgraph/tutorials/overview/): Guided examples on getting started with LangGraph.\n* [LangChain Academy](https://academy.langchain.com/courses/intro-to-langgraph): Learn the basics of LangGraph in our free, structured course.\n* [Templates](https://langchain-ai.github.io/langgraph/concepts/template_applications/): Pre-built reference apps for common agentic workflows (e.g. ReAct agent, memory, retrieval etc.) that can be cloned and adapted.\n* [Case studies](https://www.langchain.com/built-with-langgraph): Hear how industry leaders use LangGraph to ship AI applications at scale.\n\nAcknowledgements\n----------------\n\n[](https://github.com/langchain-ai/langgraph#acknowledgements)\n\nLangGraph is inspired by [Pregel](https://research.google/pubs/pub37252/) and [Apache Beam](https://beam.apache.org/). The public interface draws inspiration from [NetworkX](https://networkx.org/documentation/latest/). LangGraph is built by LangChain Inc, the creators of LangChain, but can be used without LangChain.\n\nAbout\n-----\n\nBuild resilient language agents as graphs.\n\n[langchain-ai.github.io/langgraph/](https://langchain-ai.github.io/langgraph/ \"https://langchain-ai.github.io/langgraph/\")\n\n### Resources\n\n[Readme](https://github.com/langchain-ai/langgraph#readme-ov-file)\n\n### License\n\n[MIT license](https://github.com/langchain-ai/langgraph#MIT-1-ov-file)\n\n### Security policy\n\n[Security policy](https://github.com/langchain-ai/langgraph#security-ov-file)\n\n### Uh oh!\n\nThere was an error while loading. Please reload this page.\n\n[Activity](https://github.com/langchain-ai/langgraph/activity)\n\n[Custom properties](https://github.com/langchain-ai/langgraph/custom-properties)\n\n### Stars\n\n[**14.5k** stars](https://github.com/langchain-ai/langgraph/stargazers)\n\n### Watchers\n\n[**95** watching](https://github.com/langchain-ai/langgraph/watchers)\n\n### Forks\n\n[**2.4k** forks](https://github.com/langchain-ai/langgraph/forks)\n\n[Report repository](https://github.com/contact/report-content?content_url=https%3A%2F%2Fgithub.com%2Flangchain-ai%2Flanggraph&report=langchain-ai+%28user%29)\n\n[Releases 343](https://github.com/langchain-ai/langgraph/releases)\n------------------------------------------------------------------\n\n[checkpoint==2.1.0 Latest Jun 16, 2025](https://github.com/langchain-ai/langgraph/releases/tag/checkpoint%3D%3D2.1.0)\n\n[+ 342 releases](https://github.com/langchain-ai/langgraph/releases)\n\n[Packages 0](https://github.com/orgs/langchain-ai/packages?repo_name=langgraph)\n-------------------------------------------------------------------------------\n\n No packages published \n\n[Used by 31.2k](https://github.com/langchain-ai/langgraph/network/dependents)\n-----------------------------------------------------------------------------\n\n[*  *  *  *  *  *  *  *  + 31,173](https://github.com/langchain-ai/langgraph/network/dependents)\n\n[Contributors 206](https://github.com/langchain-ai/langgraph/graphs/contributors)\n---------------------------------------------------------------------------------\n\n* [](https://github.com/nfcampos)\n* [](https://github.com/hinthornw)\n* [](https://github.com/dqbd)\n* [](https://github.com/eyurtsev)\n* [](https://github.com/andrewnguonly)\n* [](https://github.com/hwchase17)\n* [](https://github.com/sydney-runkle)\n* [](https://github.com/isahers1)\n* [](https://github.com/rlancemartin)\n* [](https://github.com/ccurme)\n* [](https://github.com/bracesproul)\n* [](https://github.com/lc-arjun)\n* [](https://github.com/lnhsingh)\n* [](https://github.com/jacoblee93)\n\n[+ 192 contributors](https://github.com/langchain-ai/langgraph/graphs/contributors)\n\nLanguages\n---------\n\n* [Python 94.1%](https://github.com/langchain-ai/langgraph/search?l=python)\n* [TypeScript 5.4%](https://github.com/langchain-ai/langgraph/search?l=typescript)\n* Other 0.5%\n\nFooter\n------\n\n[](https://github.com/) Β© 2025 GitHub,Inc. \n\n### Footer navigation\n\n* [Terms](https://docs.github.com/site-policy/github-terms/github-terms-of-service)\n* [Privacy](https://docs.github.com/site-policy/privacy-policies/github-privacy-statement)\n* [Security](https://github.com/security)\n* [Status](https://www.githubstatus.com/)\n* [Docs](https://docs.github.com/)\n* [Contact](https://support.github.com/?tags=dotcom-footer)\n* Manage cookies \n* Do not share my personal information \n\n You canβt perform that action at this time.\n"}], "response_time": 1.63}
+2025-07-31 18:59:57 INFO: +++ Tool Call +++
+{"query": "social media mentions https://github.com/hwchase17/deepagents", "follow_up_questions": null, "answer": null, "images": [], "results": [{"url": "https://github.com/langchain-ai/social-media-agent", "title": "GitHub - langchain-ai/social-media-agent", "content": "Missing: mentions hwchase17/ deepagents", "score": 0.81347597, "raw_content": "## Navigation Menu\n\n# Search code, repositories, users, issues, pull requests...\n\n# Provide feedback\n\nWe read every piece of feedback, and take your input very seriously.\n\n# Saved searches\n\n## Use saved searches to filter your results more quickly\n\nTo see all available qualifiers, see our [documentation](https://docs.github.com/search-github/github-code-search/understanding-github-code-search-syntax).\n\nπ² An agent for sourcing, curating, and scheduling social media posts with human-in-the-loop.\n\n### License\n\n### Uh oh!\n\nThere was an error while loading. Please reload this page.\n\nThere was an error while loading. Please reload this page.\n\n# langchain-ai/social-media-agent\n\n## Folders and files\n\n| Name | | Name | Last commit message | Last commit date |\n| --- | --- | --- | --- | --- |\n| Latest commit History[539 Commits](/langchain-ai/social-media-agent/commits/main/) | | |\n| [.github/workflows](/langchain-ai/social-media-agent/tree/main/.github/workflows \"This path skips through empty directories\") | | [.github/workflows](/langchain-ai/social-media-agent/tree/main/.github/workflows \"This path skips through empty directories\") | | |\n| [.vscode](/langchain-ai/social-media-agent/tree/main/.vscode \".vscode\") | | [.vscode](/langchain-ai/social-media-agent/tree/main/.vscode \".vscode\") | | |\n| [memory-v2](/langchain-ai/social-media-agent/tree/main/memory-v2 \"memory-v2\") | | [memory-v2](/langchain-ai/social-media-agent/tree/main/memory-v2 \"memory-v2\") | | |\n| [scripts](/langchain-ai/social-media-agent/tree/main/scripts \"scripts\") | | [scripts](/langchain-ai/social-media-agent/tree/main/scripts \"scripts\") | | |\n| [slack-messaging](/langchain-ai/social-media-agent/tree/main/slack-messaging \"slack-messaging\") | | [slack-messaging](/langchain-ai/social-media-agent/tree/main/slack-messaging \"slack-messaging\") | | |\n| [src](/langchain-ai/social-media-agent/tree/main/src \"src\") | | [src](/langchain-ai/social-media-agent/tree/main/src \"src\") | | |\n| [static](/langchain-ai/social-media-agent/tree/main/static \"static\") | | [static](/langchain-ai/social-media-agent/tree/main/static \"static\") | | |\n| [.codespellignore](/langchain-ai/social-media-agent/blob/main/.codespellignore \".codespellignore\") | | [.codespellignore](/langchain-ai/social-media-agent/blob/main/.codespellignore \".codespellignore\") | | |\n| [.dockerignore](/langchain-ai/social-media-agent/blob/main/.dockerignore \".dockerignore\") | | [.dockerignore](/langchain-ai/social-media-agent/blob/main/.dockerignore \".dockerignore\") | | |\n| [.editorconfig](/langchain-ai/social-media-agent/blob/main/.editorconfig \".editorconfig\") | | [.editorconfig](/langchain-ai/social-media-agent/blob/main/.editorconfig \".editorconfig\") | | |\n| [.env.full.example](/langchain-ai/social-media-agent/blob/main/.env.full.example \".env.full.example\") | | [.env.full.example](/langchain-ai/social-media-agent/blob/main/.env.full.example \".env.full.example\") | | |\n| [.env.quickstart.example](/langchain-ai/social-media-agent/blob/main/.env.quickstart.example \".env.quickstart.example\") | | [.env.quickstart.example](/langchain-ai/social-media-agent/blob/main/.env.quickstart.example \".env.quickstart.example\") | | |\n| [.eslintrc.cjs](/langchain-ai/social-media-agent/blob/main/.eslintrc.cjs \".eslintrc.cjs\") | | [.eslintrc.cjs](/langchain-ai/social-media-agent/blob/main/.eslintrc.cjs \".eslintrc.cjs\") | | |\n| [.gitignore](/langchain-ai/social-media-agent/blob/main/.gitignore \".gitignore\") | | [.gitignore](/langchain-ai/social-media-agent/blob/main/.gitignore \".gitignore\") | | |\n| [FEATURES.md](/langchain-ai/social-media-agent/blob/main/FEATURES.md \"FEATURES.md\") | | [FEATURES.md](/langchain-ai/social-media-agent/blob/main/FEATURES.md \"FEATURES.md\") | | |\n| [LICENSE](/langchain-ai/social-media-agent/blob/main/LICENSE \"LICENSE\") | | [LICENSE](/langchain-ai/social-media-agent/blob/main/LICENSE \"LICENSE\") | | |\n| [README.md](/langchain-ai/social-media-agent/blob/main/README.md \"README.md\") | | [README.md](/langchain-ai/social-media-agent/blob/main/README.md \"README.md\") | | |\n| [jest.config.js](/langchain-ai/social-media-agent/blob/main/jest.config.js \"jest.config.js\") | | [jest.config.js](/langchain-ai/social-media-agent/blob/main/jest.config.js \"jest.config.js\") | | |\n| [jest.setup.cjs](/langchain-ai/social-media-agent/blob/main/jest.setup.cjs \"jest.setup.cjs\") | | [jest.setup.cjs](/langchain-ai/social-media-agent/blob/main/jest.setup.cjs \"jest.setup.cjs\") | | |\n| [langgraph.json](/langchain-ai/social-media-agent/blob/main/langgraph.json \"langgraph.json\") | | [langgraph.json](/langchain-ai/social-media-agent/blob/main/langgraph.json \"langgraph.json\") | | |\n| [package.json](/langchain-ai/social-media-agent/blob/main/package.json \"package.json\") | | [package.json](/langchain-ai/social-media-agent/blob/main/package.json \"package.json\") | | |\n| [pyproject.toml](/langchain-ai/social-media-agent/blob/main/pyproject.toml \"pyproject.toml\") | | [pyproject.toml](/langchain-ai/social-media-agent/blob/main/pyproject.toml \"pyproject.toml\") | | |\n| [tsconfig.json](/langchain-ai/social-media-agent/blob/main/tsconfig.json \"tsconfig.json\") | | [tsconfig.json](/langchain-ai/social-media-agent/blob/main/tsconfig.json \"tsconfig.json\") | | |\n| [uv.lock](/langchain-ai/social-media-agent/blob/main/uv.lock \"uv.lock\") | | [uv.lock](/langchain-ai/social-media-agent/blob/main/uv.lock \"uv.lock\") | | |\n| [yarn.lock](/langchain-ai/social-media-agent/blob/main/yarn.lock \"yarn.lock\") | | [yarn.lock](/langchain-ai/social-media-agent/blob/main/yarn.lock \"yarn.lock\") | | |\n| View all files | | |\n\n## Latest commit\n\n## History\n\n## Repository files navigation\n\n# Social Media Agent\n\nThis repository contains an 'agent' which can take in a URL, and generate a Twitter & LinkedIn post based on the content of the URL. It uses a human-in-the-loop (HITL) flow to handle authentication with different social media platforms, and to allow the user to make changes, or accept/reject the generated post.\n\n[](/langchain-ai/social-media-agent/blob/main/static/agent_flow.png)\n\n\n\n## Table of contents\n\n# Quickstart\n\nTip\n\nπ₯ For a visual guide, check out our [step-by-step video tutorial](https://youtu.be/TmTl5FMgkCQ) that walks you through the account setup process and project configuration.\n\nThis quickstart covers how to setup the Social Media Agent in a basic setup mode. This is the quickest way to get up and running, however it will lack some of the features of the full setup mode. See [here](#advanced-setup) for the full setup guide.\n\nTo get started, you'll need the following API keys/software:\n\n## Setup Instructions\n\n### Clone the repository:\n\n### Install dependencies:\n\n### Set environment variables.\n\nCopy the values of the quickstart `.env.quickstart.example` to `.env`, then add the values:\n\n`.env.quickstart.example`\n`.env`\n\nOnce done, ensure you have the following environment variables set:\n\nIf you plan to post to LinkedIn as an organization (rather than as yourself), you'll also need to set:\n\n### Install LangGraph CLI\n\nThen run the following command to check the CLI is installed:\n\nClick [here](https://langchain-ai.github.io/langgraph/cloud/reference/cli/) to read the full download instructions for the LangGraph CLI.\n\n### Start the LangGraph server:\n\nTo start the LangGraph server, run this script:\n\nUnder the hood, this will execute the following command:\n\nNote\n\nThe first time running this command (or if a new version of `@langchain/langgraph-cli` has been released), it will ask you to accept an install for the CLI. Enter `y` to accept.\n\n`@langchain/langgraph-cli`\n`y`\n\nOnce the server is ready, you can execute the following command to generate a post:\n\nYou may also modify this script to pass different URLs to generate posts for other content.\n\nThis will kick off a new run to generate a post on a [LangChain blog post](https://blog.langchain.dev/customers-appfolio/).\n\nTo view the output, either inspect it in LangSmith, or use Agent Inbox.\n\nTip\n\nFollow these steps to setup & configure the Agent Inbox: [Setup Agent Inbox Guide](#setup-agent-inbox)\n\n# Advanced Setup\n\n[](/langchain-ai/social-media-agent/blob/main/static/graph_screenshot.png)\n\n\n\nTo use all of the features of the Social Media Agent, you'll need the following:\n\n## Setup Instructions\n\n### Clone the repository:\n\n### Install dependencies:\n\n### Set environment variables.\n\nCopy the values of the full env example file `.env.full.example` to `.env`, then update the values as needed.\n\n`.env.full.example`\n`.env`\n\n### Setup authentication\n\nThe agent needs your authorization to read and write to social media platforms. There are two ways to authorize the agent:\n\nYou can use either method, but not both.\n\n#### Arcade setup\n\nCreate an Arcade account [here](https://www.arcade.dev). After you register, [get an Arcade API key](https://docs.arcade.dev/home/quickstart?lang=typescript). Set this value as `ARCADE_API_KEY` in your `.env` file.\n\n`ARCADE_API_KEY`\n`.env`\n\nThen, you will need to set these environment variables in your `.env` file:\n\n`.env`\n`TWITTER_USER_ID`\n`LINKEDIN_USER_ID`\n\nMake sure you have the `USE_ARCADE_AUTH` environment variable set to `true` to have the graph use Arcade authentication.\n\n`USE_ARCADE_AUTH`\n`true`\n\nIf you plan to post to LinkedIn as an organization (rather than as yourself), you'll also need to set:\n\nNote\n\nIf you want to upload media to Twitter, you will still need to set up your own Twitter developer account (below) in addition to using Arcade.\n\nIf you are only planning to read/write text posts on Twitter, you can use Arcade without any additional setup.\n\n#### Twitter app setup\n\nYou'll need to follow these instructions if you plan on uploading media to Twitter, and/or you are not using Arcade for authorization.\n\n`API Key`\n`API Key Secret`\n`Bearer Token`\n`TWITTER_API_KEY`\n`TWITTER_API_KEY_SECRET`\n`TWITTER_BEARER_TOKEN`\n`.env`\n`User authentication settings`\n`Set up`\n`App permissions`\n`Read and write`\n`Type of App`\n`Web App, Automated App or Bot`\n`App info`\n`Callback URI/Redirect URL`\n`http://localhost:3000/auth/twitter/callback`\n`Website URL`\n`Client ID`\n`Client Secret`\n`TWITTER_CLIENT_ID`\n`TWITTER_CLIENT_SECRET`\n`.env`\n\nOnce done, run the `yarn start:auth` command to run the Twitter OAuth server. Open in your browser, and click `Login with Twitter`.\n\n`yarn start:auth`\n`Login with Twitter`\n\nAfter authorizing your account with the app, navigate to your terminal where you'll see a JSON object logged. Copy the `token` and `tokenSecret` values and set them as `TWITTER_USER_TOKEN` and `TWITTER_USER_TOKEN_SECRET` in your `.env` file.\n\n`token`\n`tokenSecret`\n`TWITTER_USER_TOKEN`\n`TWITTER_USER_TOKEN_SECRET`\n`.env`\n\n#### LinkedIn app setup\n\nYou'll need to follow these instructions if you plan on posting to LinkedIn and are not using Arcade for authorization.\n\n`Auth`\n`http://localhost:3000/auth/linkedin/callback`\n`Products`\n`Share on LinkedIn`\n`Sign In with LinkedIn using OpenID Connect`\n`Advertising API`\n`ADMINISTRATOR`\n`DIRECT_SPONSORED_CONTENT_POSTER`\n`RECRUITING_POSTER`\n`Settings`\n`Verify`\n`w_organization_social`\n`/auth/linkedin`\n`openid profile email w_member_social w_organization_social`\n`LINKEDIN_ORGANIZATION_ID`\n`https://www.linkedin.com/company/12345678/admin/dashboard/`\n`12345678`\n\n[!NOTE]\nIf you plan on only posting from the company account, you can set the `POST_TO_LINKEDIN_ORGANIZATION` to `\"true\"` in your `.env` file. If you want to choose dynamically, you can set this to `true`/`false` in the configurable fields (`postToLinkedInOrganization`) when invoking the `generate_post` graph.\n\n`POST_TO_LINKEDIN_ORGANIZATION`\n`\"true\"`\n`.env`\n`true`\n`false`\n`postToLinkedInOrganization`\n`generate_post`\n\nThis value will take precedence over the `POST_TO_LINKEDIN_ORGANIZATION` environment variable.\n\n`POST_TO_LINKEDIN_ORGANIZATION`\n`.env`\n`LINKEDIN_CLIENT_ID`\n`LINKEDIN_CLIENT_SECRET`\n`yarn start:auth`\n`Login with LinkedIn`\n`access_token`\n`sub`\n`LINKEDIN_ACCESS_TOKEN`\n`access_token`\n`LINKEDIN_PERSON_URN`\n`sub`\n`.env`\n\n### Setup Supabase\n\nSupabase is required for storing images found/generated by the agent. This step is not required for running the agent in basic setup mode.\n\nTo setup Supabase, create an account and a new project.\n\nSet the `SUPABASE_URL` and `SUPABASE_SERVICE_ROLE_KEY` environment variables to the values provided by Supabase.\n\n`SUPABASE_URL`\n`SUPABASE_SERVICE_ROLE_KEY`\n\nCreate a new storage bucket called `images`. Make sure the bucket is set to public to the image URLs are accessible. Also ensure the max upload size is set to at least 5MB inside the global project settings, and the bucket specific settings.\n\n`images`\n\n### Setup Slack\n\nSlack integration is optional, but recommended if you intend on using the `ingest_data` agent, or having updates sent to Slack.\n\n`ingest_data`\n\nThis agent can be used in a cron job to fetch messages from a Slack channel, and call the `generate_post` graph for each message. We use this flow internally at LangChain to enable having a single Slack channel for submitting relevant URLs to the agent, which are then turned into posts once daily.\n\n`generate_post`\n\nTo configure the Slack integration, create a new Slack app and install it into your desired Slack workspace.\n\nOnce installed, ensure it has access to the channel you want to ingest messages from. Additionally, if you want it to send update messages to Slack, it will need write permissions to the workspace.\n\nFinally, make sure the `SLACK_BOT_TOKEN` environment variable is set in your `.env` file. Then, when you create a cron (see the [Setup Crons](#setup-crons) section), you'll only have to pass in the channel ID to ingest messages from.\n\n`SLACK_BOT_TOKEN`\n`.env`\n\nTo enable sending updates to Slack, add a `SLACK_CHANNEL_ID` environment variable to your `.env` file with the channel ID you want to send updates to (at LangChain, we have one channel for sending content links, and a separate one for sending update messages).\n\n`SLACK_CHANNEL_ID`\n`.env`\n\n### Setup GitHub\n\nThe GitHub API token is required to fetch details about GitHub repository URLs submitted to the agent. This is not required if you do not plan on sending GitHub URLs to the agent.\n\nTo get a GitHub API token, create a new fine grained token with the `Public Repositories (read-only)` scope at a minimum. If you intend on using this agent for private GitHub repositories, you'll need to give the token access to those repositories as well.\n\n`Public Repositories (read-only)`\n\nEnsure this is set as `GITHUB_TOKEN` in your `.env` file.\n\n`GITHUB_TOKEN`\n`.env`\n\n# Usage\n\n## Generate Post\n\nOnce all the setup steps have been completed, start your graph server by running:\n\nNote\n\nThe first time running this command (or if a new version of `@langchain/langgraph-cli` has been released), it will ask you to accept an install for the CLI. Enter `y` to accept.\n\n`@langchain/langgraph-cli`\n`y`\n\nOnce the server is ready, you can execute the following command to generate a post:\n\n(before doing this, you should edit the file so that the text only mode is set to false: `[TEXT_ONLY_MODE]: false` if using advanced setup mode)\n\n`[TEXT_ONLY_MODE]: false`\n\nThis will kick off a new run to generate a post on a [LangChain blog post](https://blog.langchain.dev/customers-appfolio/).\n\nTo view the output, either inspect it in LangSmith, or use [the Agent Inbox](#setup-agent-inbox).\n\nYou may also modify this script to pass different URLs to generate posts for other content.\n\n## Setup Crons\n\nThe Agent Inbox is most powerful when used with cron jobs. Doing this allows you to send links to generate posts on to a Slack channel, and have the cron job check for these links asynchronously. With this setup, you can send links to Slack, and have the inbox handle the rest while you sleep.\n\nWe have a series of scripts to help with this, which you can find in the [`scripts/crons`](/langchain-ai/social-media-agent/blob/main/scripts/crons) directory. In this section, we'll explain how to quickly create a cron for ingesting links to your graph from Slack.\n\n`scripts/crons`\n\n### Slack Setup\n\nBefore we get started, ensure you have the proper Slack integration setup, as described in the [Setup Slack](#setup-slack) section. Then, edit the [`create-cron.ts`](/langchain-ai/social-media-agent/blob/main/scripts/crons/create-cron.ts) file, and set the `slackChannelId` field to the channel ID of the channel you want to ingest links from.\n\n`create-cron.ts`\n`slackChannelId`\n\nAfter editing, run the following command to create the cron:\n\nThis will create a new cron job that will ingest links from Slack into your graph, once daily.\n\n## Prebuilt Scripts\n\nFor more information on all of the prebuilt scripts, see the [`scripts/README.md`](/langchain-ai/social-media-agent/blob/main/scripts/README.md) file.\n\n`scripts/README.md`\n\n# Setup Agent Inbox\n\nThe Agent Inbox is the easiest way to view interrupted events, and manage accepting, responding, or other allowed actions. To view your events in the inbox, you can either add your graph to the deployed version of the Agent Inbox, or clone & run the Agent Inbox locally.\n\n## Using the deployed inbox\n\nThe Agent Inbox is setup in a way that allows for any graph --local or deployed-- to be added & accessed via the UI.\n\nTo add your local graph to the inbox, visit the deployed site here: [dev.agentinbox.ai](https://dev.agentinbox.ai/).\n\nIf it's your first time vising the site, you'll immediately be prompted to add a new graph. Fill out the form with the following values:\n\n`generate_post`\n`http://localhost:54367`\n`Generate Post (local)`\n\nAfter saving, you should be able to view your graph in the inbox. If you do this after invoking your graph (and waiting for the thread to interrupt), it will automatically fetch the interrupted event.\n\n## Using the local inbox\n\nTo run the Agent Inbox locally, follow the setup instructions [here](https://github.com/langchain-ai/agent-inbox/blob/main/README.md).\n\nOnce the web server is running, open your browser and visit . This will then prompt you to add your graph to the inbox.\n\nFill out the form with the following values:\n\n`generate_post`\n`http://localhost:54367`\n`Generate Post (local)`\n\n## Using the Agent Inbox with a graph deployed on LangGraph Platform\n\nThe Agent Inbox can also be used with graph's deployed in production on LangGraph Platform. To use these graphs, the setup steps are the exact same, with the only difference being the graph API URL should be the URL of the deployed graph, and you're required to set a LangSmith API key. This is required to fetch & invoke the deployed graph. LangSmith API keys are stored in your browser's local storage, and never stored on the server.\n\n# Customization\n\n## Prompts\n\nThis agent is setup to generate posts for LangChain, using LangChain products as context. To use the agent for your own use case, you should update the following prompts/prompt sections inside the [`prompts`](/langchain-ai/social-media-agent/blob/main/src/agents/generate-post/prompts/index.ts) folder:\n\n`prompts`\n`BUSINESS_CONTEXT`\n`TWEET_EXAMPLES`\n`prompts/examples.ts`\n`POST_STRUCTURE_INSTRUCTIONS`\n`POST_CONTENT_RULES`\n\nThe prompt for the marketing report is located in the [`generate-post/nodes/generate-report/prompts.ts`](/langchain-ai/social-media-agent/blob/main/src/agents/generate-post/nodes/generate-report/prompts.ts) file. You likely don't need to update this, as it's already structured to be general.\n\n`generate-post/nodes/generate-report/prompts.ts`\n\n## Post Style\n\nThere are two main prompts to modify to change the style of the posts.\n\n`POST_STRUCTURE_INSTRUCTIONS`\n`Header`\n`Body`\n`Call to action`\n`TWEET_EXAMPLES`\n`POST_CONTENT_RULES`\n`TWEET_EXAMPLES`\n`BUSINESS_CONTEXT`\n\n## About\n\nπ² An agent for sourcing, curating, and scheduling social media posts with human-in-the-loop.\n\n### Resources\n\n### License\n\n### Uh oh!\n\nThere was an error while loading. Please reload this page.\n\nThere was an error while loading. Please reload this page.\n\n### Stars\n\n### Watchers\n\n### Forks\n\n## [Releases](/langchain-ai/social-media-agent/releases)\n\n## [Packages 0](/orgs/langchain-ai/packages?repo_name=social-media-agent)\n\n### Uh oh!\n\nThere was an error while loading. Please reload this page.\n\nThere was an error while loading. Please reload this page.\n\n## [Contributors 2](/langchain-ai/social-media-agent/graphs/contributors)\n\n### Uh oh!\n\nThere was an error while loading. Please reload this page.\n\nThere was an error while loading. Please reload this page.\n\n## Languages\n\n## Footer\n\n### Footer navigation"}, {"url": "https://github.com/hwchase17/deepagents", "title": "hwchase17/deepagents - GitHub", "content": "Missing: social media mentions", "score": 0.75312227, "raw_content": "## Navigation Menu\n\n# Search code, repositories, users, issues, pull requests...\n\n# Provide feedback\n\nWe read every piece of feedback, and take your input very seriously.\n\n# Saved searches\n\n## Use saved searches to filter your results more quickly\n\nTo see all available qualifiers, see our [documentation](https://docs.github.com/search-github/github-code-search/understanding-github-code-search-syntax).\n\n### License\n\n### Uh oh!\n\nThere was an error while loading. Please reload this page.\n\nThere was an error while loading. Please reload this page.\n\n# hwchase17/deepagents\n\n## Folders and files\n\n| Name | | Name | Last commit message | Last commit date |\n| --- | --- | --- | --- | --- |\n| Latest commit History[8 Commits](/hwchase17/deepagents/commits/master/) | | |\n| [examples/research](/hwchase17/deepagents/tree/master/examples/research \"This path skips through empty directories\") | | [examples/research](/hwchase17/deepagents/tree/master/examples/research \"This path skips through empty directories\") | | |\n| [src/deepagents](/hwchase17/deepagents/tree/master/src/deepagents \"This path skips through empty directories\") | | [src/deepagents](/hwchase17/deepagents/tree/master/src/deepagents \"This path skips through empty directories\") | | |\n| [.gitignore](/hwchase17/deepagents/blob/master/.gitignore \".gitignore\") | | [.gitignore](/hwchase17/deepagents/blob/master/.gitignore \".gitignore\") | | |\n| [LICENSE](/hwchase17/deepagents/blob/master/LICENSE \"LICENSE\") | | [LICENSE](/hwchase17/deepagents/blob/master/LICENSE \"LICENSE\") | | |\n| [README.md](/hwchase17/deepagents/blob/master/README.md \"README.md\") | | [README.md](/hwchase17/deepagents/blob/master/README.md \"README.md\") | | |\n| [deep\\_agents.png](/hwchase17/deepagents/blob/master/deep_agents.png \"deep_agents.png\") | | [deep\\_agents.png](/hwchase17/deepagents/blob/master/deep_agents.png \"deep_agents.png\") | | |\n| [pyproject.toml](/hwchase17/deepagents/blob/master/pyproject.toml \"pyproject.toml\") | | [pyproject.toml](/hwchase17/deepagents/blob/master/pyproject.toml \"pyproject.toml\") | | |\n| View all files | | |\n\n## Latest commit\n\n## History\n\n## Repository files navigation\n\n# π§ π€Deep Agents\n\nUsing an LLM to call tools in a loop is the simplest form of an agent.\nThis architecture, however, can yield agents that are βshallowβ and fail to plan and act over longer, more complex tasks.\nApplications like βDeep Researchβ, \"Manus\", and βClaude Codeβ have gotten around this limitation by implementing a combination of four things:\na **planning tool**, **sub agents**, access to a **file system**, and a **detailed prompt**.\n\n[](/hwchase17/deepagents/blob/master/deep_agents.png)\n\n\n\n`deepagents` is a Python package that implements these in a general purpose way so that you can easily create a Deep Agent for your application.\n\n`deepagents`\n\n**Acknowledgements: This project was primarily inspired by Claude Code, and initially was largely an attempt to see what made Claude Code general purpose, and make it even more so.**\n\n## Installation\n\n## Usage\n\n(To run the example below, will need to `pip install tavily-python`)\n\n`pip install tavily-python`\n\nSee [examples/research/research\\_agent.py](/hwchase17/deepagents/blob/master/examples/research/research_agent.py) for a more complex example.\n\nThe agent created with `create_deep_agent` is just a LangGraph graph - so you can interact with it (streaming, human-in-the-loop, memory, studio)\nin the same way you would any LangGraph agent.\n\n`create_deep_agent`\n\n## Creating a custom deep agent\n\nThere are three parameters you can pass to `create_deep_agent` to create your own custom deep agent.\n\n`create_deep_agent`\n\n### `tools` (Required)\n\n`tools`\n\nThe first argument to `create_deep_agent` is `tools`.\nThis should be a list of functions or LangChain `@tool` objects.\nThe agent (and any subagents) will have access to these tools.\n\n`create_deep_agent`\n`tools`\n`@tool`\n\n### `instructions` (Required)\n\n`instructions`\n\nThe second argument to `create_deep_agent` is `instructions`.\nThis will serve as part of the prompt of the deep agent.\nNote that there is a [built in system prompt](#built-in-prompt) as well, so this is not the *entire* prompt the agent will see.\n\n`create_deep_agent`\n`instructions`\n\n### `subagents` (Optional)\n\n`subagents`\n\nA keyword-only argument to `create_deep_agent` is `subagents`.\nThis can be used to specify any custom subagents this deep agent will have access to.\nYou can read more about why you would want to use subagents [here](#sub-agents)\n\n`create_deep_agent`\n`subagents`\n\n`subagents` should be a list of dictionaries, where each dictionary follow this schema:\n\n`subagents`\n\nTo use it looks like:\n\n### `model` (Optional)\n\n`model`\n\nBy default, `deepagents` will use `\"claude-sonnet-4-20250514\"`. If you want to use a different model,\nyou can pass a [LangChain model object](https://python.langchain.com/docs/integrations/chat/).\n\n`deepagents`\n`\"claude-sonnet-4-20250514\"`\n\n## Deep Agent Details\n\nThe below components are built into `deepagents` and helps make it work for deep tasks off-the-shelf.\n\n`deepagents`\n\n### System Prompt\n\n`deepagents` comes with a [built-in system prompt](/hwchase17/deepagents/blob/master/src/deepagents/prompts.py). This is relatively detailed prompt that is heavily based on and inspired by [attempts](https://github.com/kn1026/cc/blob/main/claudecode.md) to [replicate](https://github.com/asgeirtj/system_prompts_leaks/blob/main/Anthropic/claude-code.md)\nClaude Code's system prompt. It was made more general purpose than Claude Code's system prompt.\nThis contains detailed instructions for how to use the built-in planning tool, file system tools, and sub agents.\nNote that part of this system prompt [can be customized](#promptprefix--required-)\n\n`deepagents`\n\nWithout this default system prompt - the agent would not be nearly as successful at going as it is.\nThe importance of prompting for creating a \"deep\" agent cannot be understated.\n\n### Planing Tool\n\n`deepagents` comes with a built-in planning tool. This planning tool is very simple and is based on ClaudeCode's TodoWrite tool.\nThis tool doesn't actually do anything - it is just a way for the agent to come up with a plan, and then have that in the context to help keep it on track.\n\n`deepagents`\n\n### File System Tools\n\n`deepagents` comes with four built-in file system tools: `ls`, `edit_file`, `read_file`, `write_file`.\nThese do not actually use a file system - rather, they mock out a file system using LangGraph's State object.\nThis means you can easily run many of these agents on the same machine without worrying that they will edit the same underlying files.\n\n`deepagents`\n`ls`\n`edit_file`\n`read_file`\n`write_file`\n\nRight now the \"file system\" will only be one level deep (no sub directories).\n\nThese files can be passed in (and also retrieved) by using the `files` key in the LangGraph State object.\n\n`files`\n\n### Sub Agents\n\n`deepagents` comes with the built-in ability to call sub agents (based on Claude Code).\nIt has access to a `general-purpose` subagent at all times - this is a subagent with the same instructions as the main agent and all the tools that is has access to.\nYou can also specify [custom sub agents](#subagents--optional-) with their own instructions and tools.\n\n`deepagents`\n`general-purpose`\n\nSub agents are useful for [\"context quarantine\"](https://www.dbreunig.com/2025/06/26/how-to-fix-your-context.html#context-quarantine) (to help not pollute the overall context of the main agent)\nas well as custom instructions.\n\n## Roadmap\n\n[] Allow users to customize full system prompt\n[] Code cleanliness (type hinting, docstrings, formating)\n[] Allow for more of a robust virtual filesystem\n[] Create an example of a deep coding agent built on top of this\n[] Benchmark the example of [deep research agent](/hwchase17/deepagents/blob/master/examples/research/research_agent.py)\n[] Add human-in-the-loop support for tools\n\n## About\n\n### Resources\n\n### License\n\n### Uh oh!\n\nThere was an error while loading. Please reload this page.\n\nThere was an error while loading. Please reload this page.\n\n### Stars\n\n### Watchers\n\n### Forks\n\n## [Releases](/hwchase17/deepagents/releases)\n\n## [Packages 0](/users/hwchase17/packages?repo_name=deepagents)\n\n## Languages\n\n## Footer\n\n### Footer navigation"}], "response_time": 2.09}
+2025-07-31 18:59:57 INFO: +++ Tool Call +++
+{"query": "README.md https://github.com/hwchase17/deepagents", "follow_up_questions": null, "answer": null, "images": [], "results": [{"url": "https://github.com/hwchase17/deepagents", "title": "hwchase17/deepagents - GitHub", "content": "deepagents is a Python package that implements these in a general purpose way so that you can easily create a Deep Agent for your application.", "score": 0.78413767, "raw_content": "## Navigation Menu\n\n# Search code, repositories, users, issues, pull requests...\n\n# Provide feedback\n\nWe read every piece of feedback, and take your input very seriously.\n\n# Saved searches\n\n## Use saved searches to filter your results more quickly\n\nTo see all available qualifiers, see our [documentation](https://docs.github.com/search-github/github-code-search/understanding-github-code-search-syntax).\n\n### License\n\n### Uh oh!\n\nThere was an error while loading. Please reload this page.\n\nThere was an error while loading. Please reload this page.\n\n# hwchase17/deepagents\n\n## Folders and files\n\n| Name | | Name | Last commit message | Last commit date |\n| --- | --- | --- | --- | --- |\n| Latest commit History[8 Commits](/hwchase17/deepagents/commits/master/) | | |\n| [examples/research](/hwchase17/deepagents/tree/master/examples/research \"This path skips through empty directories\") | | [examples/research](/hwchase17/deepagents/tree/master/examples/research \"This path skips through empty directories\") | | |\n| [src/deepagents](/hwchase17/deepagents/tree/master/src/deepagents \"This path skips through empty directories\") | | [src/deepagents](/hwchase17/deepagents/tree/master/src/deepagents \"This path skips through empty directories\") | | |\n| [.gitignore](/hwchase17/deepagents/blob/master/.gitignore \".gitignore\") | | [.gitignore](/hwchase17/deepagents/blob/master/.gitignore \".gitignore\") | | |\n| [LICENSE](/hwchase17/deepagents/blob/master/LICENSE \"LICENSE\") | | [LICENSE](/hwchase17/deepagents/blob/master/LICENSE \"LICENSE\") | | |\n| [README.md](/hwchase17/deepagents/blob/master/README.md \"README.md\") | | [README.md](/hwchase17/deepagents/blob/master/README.md \"README.md\") | | |\n| [deep\\_agents.png](/hwchase17/deepagents/blob/master/deep_agents.png \"deep_agents.png\") | | [deep\\_agents.png](/hwchase17/deepagents/blob/master/deep_agents.png \"deep_agents.png\") | | |\n| [pyproject.toml](/hwchase17/deepagents/blob/master/pyproject.toml \"pyproject.toml\") | | [pyproject.toml](/hwchase17/deepagents/blob/master/pyproject.toml \"pyproject.toml\") | | |\n| View all files | | |\n\n## Latest commit\n\n## History\n\n## Repository files navigation\n\n# π§ π€Deep Agents\n\nUsing an LLM to call tools in a loop is the simplest form of an agent.\nThis architecture, however, can yield agents that are βshallowβ and fail to plan and act over longer, more complex tasks.\nApplications like βDeep Researchβ, \"Manus\", and βClaude Codeβ have gotten around this limitation by implementing a combination of four things:\na **planning tool**, **sub agents**, access to a **file system**, and a **detailed prompt**.\n\n[](/hwchase17/deepagents/blob/master/deep_agents.png)\n\n\n\n`deepagents` is a Python package that implements these in a general purpose way so that you can easily create a Deep Agent for your application.\n\n`deepagents`\n\n**Acknowledgements: This project was primarily inspired by Claude Code, and initially was largely an attempt to see what made Claude Code general purpose, and make it even more so.**\n\n## Installation\n\n## Usage\n\n(To run the example below, will need to `pip install tavily-python`)\n\n`pip install tavily-python`\n\nSee [examples/research/research\\_agent.py](/hwchase17/deepagents/blob/master/examples/research/research_agent.py) for a more complex example.\n\nThe agent created with `create_deep_agent` is just a LangGraph graph - so you can interact with it (streaming, human-in-the-loop, memory, studio)\nin the same way you would any LangGraph agent.\n\n`create_deep_agent`\n\n## Creating a custom deep agent\n\nThere are three parameters you can pass to `create_deep_agent` to create your own custom deep agent.\n\n`create_deep_agent`\n\n### `tools` (Required)\n\n`tools`\n\nThe first argument to `create_deep_agent` is `tools`.\nThis should be a list of functions or LangChain `@tool` objects.\nThe agent (and any subagents) will have access to these tools.\n\n`create_deep_agent`\n`tools`\n`@tool`\n\n### `instructions` (Required)\n\n`instructions`\n\nThe second argument to `create_deep_agent` is `instructions`.\nThis will serve as part of the prompt of the deep agent.\nNote that there is a [built in system prompt](#built-in-prompt) as well, so this is not the *entire* prompt the agent will see.\n\n`create_deep_agent`\n`instructions`\n\n### `subagents` (Optional)\n\n`subagents`\n\nA keyword-only argument to `create_deep_agent` is `subagents`.\nThis can be used to specify any custom subagents this deep agent will have access to.\nYou can read more about why you would want to use subagents [here](#sub-agents)\n\n`create_deep_agent`\n`subagents`\n\n`subagents` should be a list of dictionaries, where each dictionary follow this schema:\n\n`subagents`\n\nTo use it looks like:\n\n### `model` (Optional)\n\n`model`\n\nBy default, `deepagents` will use `\"claude-sonnet-4-20250514\"`. If you want to use a different model,\nyou can pass a [LangChain model object](https://python.langchain.com/docs/integrations/chat/).\n\n`deepagents`\n`\"claude-sonnet-4-20250514\"`\n\n## Deep Agent Details\n\nThe below components are built into `deepagents` and helps make it work for deep tasks off-the-shelf.\n\n`deepagents`\n\n### System Prompt\n\n`deepagents` comes with a [built-in system prompt](/hwchase17/deepagents/blob/master/src/deepagents/prompts.py). This is relatively detailed prompt that is heavily based on and inspired by [attempts](https://github.com/kn1026/cc/blob/main/claudecode.md) to [replicate](https://github.com/asgeirtj/system_prompts_leaks/blob/main/Anthropic/claude-code.md)\nClaude Code's system prompt. It was made more general purpose than Claude Code's system prompt.\nThis contains detailed instructions for how to use the built-in planning tool, file system tools, and sub agents.\nNote that part of this system prompt [can be customized](#promptprefix--required-)\n\n`deepagents`\n\nWithout this default system prompt - the agent would not be nearly as successful at going as it is.\nThe importance of prompting for creating a \"deep\" agent cannot be understated.\n\n### Planing Tool\n\n`deepagents` comes with a built-in planning tool. This planning tool is very simple and is based on ClaudeCode's TodoWrite tool.\nThis tool doesn't actually do anything - it is just a way for the agent to come up with a plan, and then have that in the context to help keep it on track.\n\n`deepagents`\n\n### File System Tools\n\n`deepagents` comes with four built-in file system tools: `ls`, `edit_file`, `read_file`, `write_file`.\nThese do not actually use a file system - rather, they mock out a file system using LangGraph's State object.\nThis means you can easily run many of these agents on the same machine without worrying that they will edit the same underlying files.\n\n`deepagents`\n`ls`\n`edit_file`\n`read_file`\n`write_file`\n\nRight now the \"file system\" will only be one level deep (no sub directories).\n\nThese files can be passed in (and also retrieved) by using the `files` key in the LangGraph State object.\n\n`files`\n\n### Sub Agents\n\n`deepagents` comes with the built-in ability to call sub agents (based on Claude Code).\nIt has access to a `general-purpose` subagent at all times - this is a subagent with the same instructions as the main agent and all the tools that is has access to.\nYou can also specify [custom sub agents](#subagents--optional-) with their own instructions and tools.\n\n`deepagents`\n`general-purpose`\n\nSub agents are useful for [\"context quarantine\"](https://www.dbreunig.com/2025/06/26/how-to-fix-your-context.html#context-quarantine) (to help not pollute the overall context of the main agent)\nas well as custom instructions.\n\n## Roadmap\n\n[] Allow users to customize full system prompt\n[] Code cleanliness (type hinting, docstrings, formating)\n[] Allow for more of a robust virtual filesystem\n[] Create an example of a deep coding agent built on top of this\n[] Benchmark the example of [deep research agent](/hwchase17/deepagents/blob/master/examples/research/research_agent.py)\n[] Add human-in-the-loop support for tools\n\n## About\n\n### Resources\n\n### License\n\n### Uh oh!\n\nThere was an error while loading. Please reload this page.\n\nThere was an error while loading. Please reload this page.\n\n### Stars\n\n### Watchers\n\n### Forks\n\n## [Releases](/hwchase17/deepagents/releases)\n\n## [Packages 0](/users/hwchase17/packages?repo_name=deepagents)\n\n## Languages\n\n## Footer\n\n### Footer navigation"}], "response_time": 2.76}
+2025-07-31 19:04:03 INFO: === Agent Message ===
+
+Okay, let's see. The user provided a query with some context and a response. They want me to act as a reasoning process, thinking through how to approach the problem, and then give the final answer. The query seems to be about two different projects: one called "deepagents" and another related to a social media platform with a feature called "Reels." The user might be asking about the differences between these projects or how to implement similar features.
+
+First, I need to parse the information given. The context includes code snippets and documentation for "deepagents," which is a Python package for creating deep agents using LLMs, planning tools, subagents, and file system tools. There's also a mention of a social media platform's Reels feature, which involves video processing, user interaction, and content moderation.
+
+Wait, looking back, the user's query might be about the "deepagents" project, but the initial part of the context mentions a social media platform's Reels. However, the response provided seems to focus on the "deepagents" documentation. Maybe the user is confused or there's a mix-up in the context. Alternatively, they might be asking how to apply the "deepagents" framework to a Reels feature.
+
+But the original query isn't clear. The user might have intended to ask about the "deepagents" project but included some unrelated Reels information. Alternatively, they might be comparing the two. However, since the response given is about "deepagents," I should focus on that.
+
+The user's question might be: "How can I use the deepagents framework to create a deep agent for a social media platform's Reels feature?" But since the query isn't explicitly stated, I need to infer based on the context provided.
+
+Assuming the user wants to know how to apply the "deepagents" package to a Reels feature, here's how to approach it:
+
+1. **Understand the Requirements**: Reels involve video processing, user interaction, content moderation, and possibly recommendation systems. Each of these could be handled by different agents.
+
+2. **Identify Tools and Subagents**: For video processing, tools like FFmpeg or AWS MediaConvert could be integrated. Content moderation might require a subagent with specific instructions for filtering inappropriate content. A recommendation subagent could use machine learning models to suggest videos.
+
+3. **Define Instructions**: The main agent's instructions would need to coordinate these subagents. For example, "Process user-uploaded videos using the video processing tool, moderate content with the content moderation subagent, and recommend similar videos using the recommendation subagent."
+
+4. **Implement File System Tools**: Since Reels might involve storing and retrieving video files, the file system tools in deepagents (like `write_file`, `read_file`) could be used to manage media storage.
+
+5. **Planing Tool**: The planning tool can help the agent break down tasks, such as first processing the video, then moderating it, and finally recommending similar content.
+
+6. **Customize System Prompt**: The system prompt might need to be adjusted to include specific guidelines for handling video content, user privacy, and platform policies.
+
+7. **Testing and Iteration**: After setting up the agents, testing with sample videos and iterating based on performance and user feedback would be necessary.
+
+Potential challenges include ensuring that the subagents work seamlessly together, handling large video files efficiently, and maintaining real-time processing for user interactions. Also, integrating with existing platform infrastructure and APIs would be important.
+
+In the response, the assistant should outline these steps, explain how deepagents can be applied to the Reels feature, and mention any considerations or challenges. However, since the original query isn't clear, the assistant might need to ask for clarification. But based on the provided context and response, the focus is on the deepagents framework, so the answer should elaborate on that.
+
+
+To address the integration of the **`deepagents`** framework into a **social media platform's Reels feature**, here's a structured approach based on the provided context and the capabilities of `deepagents`:
+
+---
+
+### **1. Define the Problem Scope**
+The Reels feature involves:
+- **Video Processing**: Uploading, trimming, applying filters, and encoding videos.
+- **Content Moderation**: Detecting inappropriate content (e.g., nudity, hate speech).
+- **Recommendation System**: Suggesting Reels based on user preferences or trends.
+- **User Interaction**: Handling likes, comments, and shares.
+
+Each of these can be managed by specialized **subagents** within the `deepagents` framework.
+
+---
+
+### **2. Leverage `deepagents` Components**
+#### **a. Tools Integration**
+- **Video Processing Tools**: Integrate tools like `FFmpeg` or `AWS MediaConvert` for encoding/trimming videos.
+- **Moderation Tools**: Use AI-based APIs (e.g., Google Cloud Vision, Azure Content Moderator) for content detection.
+- **Recommendation Tools**: Connect to a machine learning model (e.g., collaborative filtering, NLP-based similarity) for video recommendations.
+
+#### **b. Subagents**
+- **Video Processing Subagent**:
+ - Instructions: "Process user-uploaded videos by trimming, applying filters, and encoding to MP4 format."
+ - Tools: FFmpeg, video storage APIs.
+- **Moderation Subagent**:
+ - Instructions: "Analyze video content for nudity, hate speech, or harmful imagery. Flag or block inappropriate content."
+ - Tools: Content moderation APIs.
+- **Recommendation Subagent**:
+ - Instructions: "Analyze user behavior and video metadata to recommend similar Reels."
+ - Tools: ML models, database queries.
+
+#### **c. File System Tools**
+- Use `deepagents`'s mock file system (`ls`, `read_file`, `write_file`) to manage video storage and retrieval without affecting the actual file system. For example:
+ ```python
+ state = {"files": {"user_video_123.mp4": "base64_encoded_video_data"}}
+ ```
+
+#### **d. Planning Tool**
+- The agent can use the built-in planning tool to sequence tasks (e.g., "First process the video, then moderate it, and finally recommend similar content").
+
+---
+
+### **3. Customize the System Prompt**
+The default `deepagents` system prompt can be modified to include:
+- **Platform-Specific Rules**: "Prioritize user privacy. Do not process videos without explicit consent."
+- **Task-Specific Instructions**: "For moderation, flag videos with NSFW content for human review."
+
+---
+
+### **4. Implementation Example**
+```python
+from deepagents import create_deep_agent
+
+# Define tools
+video_tools = [FFmpegTool(), AWSMediaConvertTool()]
+moderation_tools = [GoogleVisionAPI(), AzureModerator()]
+recommendation_tools = [CollaborativeFilteringModel(), NLPSimilarityModel()]
+
+# Create subagents
+video_subagent = create_deep_agent(
+ tools=video_tools,
+ instructions="Process user-uploaded videos for Reels."
+)
+
+moderation_subagent = create_deep_agent(
+ tools=moderation_tools,
+ instructions="Moderate Reels content for policy compliance."
+)
+
+# Main agent
+reels_agent = create_deep_agent(
+ tools=[video_subagent, moderation_subagent, recommendation_tools],
+ instructions="Manage the Reels feature: process, moderate, and recommend videos.",
+ subagents=[video_subagent, moderation_subagent]
+)
+```
+
+---
+
+### **5. Challenges and Considerations**
+- **Latency**: Video processing and moderation can be slow. Use asynchronous execution or cloud-based APIs.
+- **Scalability**: Ensure subagents can handle high concurrency (e.g., using distributed task queues like Celery).
+- **Error Handling**: Implement fallbacks for failed moderation (e.g., manual review) or processing errors.
+- **User Privacy**: Ensure compliance with regulations (e.g., GDPR) when storing or analyzing video data.
+
+---
+
+### **6. Roadmap for Enhancement**
+- Add **human-in-the-loop** moderation for ambiguous content.
+- Integrate **real-time analytics** for trending Reels.
+- Support **custom subagents** for niche tasks (e.g., caption generation with LLMs).
+
+---
+
+By structuring the Reels feature with `deepagents`, the platform can achieve a modular, scalable, and maintainable system for handling complex workflows.
+2025-07-31 19:04:03 INFO: Finished streaming. Log saved to logs/agent_20250731_185726.log
diff --git a/examples/research/requirements.txt b/examples/research/requirements.txt
index 740a3b934b..2e5f9eefd6 100644
--- a/examples/research/requirements.txt
+++ b/examples/research/requirements.txt
@@ -1,3 +1,6 @@
deepagents
langgraph-cli[inmem]
tavily-python
+langchain-openai
+langchain-ollama
+langchain-mcp-adapters
\ No newline at end of file
diff --git a/examples/research/research_agent.py b/examples/research/research_agent.py
index 9b3653efa5..79548ff516 100644
--- a/examples/research/research_agent.py
+++ b/examples/research/research_agent.py
@@ -1,10 +1,43 @@
import os
+import logging
from typing import Literal
+from datetime import datetime
+from dotenv import load_dotenv
+from langchain.chat_models import init_chat_model
+from deepagents import create_deep_agent
from tavily import TavilyClient
-from deepagents import create_deep_agent, SubAgent
+load_dotenv()
+
+# βββ Prepare logs directory ββββββββββββββββββββββββββββββββββββββββββββββββββββ
+logs_dir = "logs"
+os.makedirs(logs_dir, exist_ok=True)
+
+# Generate a new filename for this run
+timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
+log_filename = os.path.join(logs_dir, f"agent_{timestamp}.log")
+
+# βββ Configure Logging βββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
+logger = logging.getLogger("agent_logger")
+logger.setLevel(logging.INFO)
+
+formatter = logging.Formatter("%(asctime)s %(levelname)s: %(message)s",
+ "%Y-%m-%d %H:%M:%S")
+
+# File handler β writes to a fresh file each run
+file_handler = logging.FileHandler(log_filename, mode="w", encoding="utf-8")
+file_handler.setLevel(logging.INFO)
+file_handler.setFormatter(formatter)
+
+# Console handler β still prints live to stdout
+console_handler = logging.StreamHandler()
+console_handler.setLevel(logging.INFO)
+console_handler.setFormatter(formatter)
+
+logger.addHandler(file_handler)
+logger.addHandler(console_handler)
# Search tool to use to do research
@@ -155,9 +188,29 @@ def internet_search(
Use this to run an internet search for a given query. You can specify the number of results, the topic, and whether raw content should be included.
"""
+
# Create the agent
agent = create_deep_agent(
[internet_search],
research_instructions,
subagents=[critique_sub_agent, research_sub_agent],
-).with_config({"recursion_limit": 1000})
+ model=init_chat_model(
+ model="ollama:qwen3:14b",
+ temperature=0.0,
+ max_tokens=40000,
+ )
+).with_config({"recursion_limit": 100})
+
+
+# Stream the agent
+result = agent.stream({"messages": [{"role": "user", "content": "Provide a summary of https://github.com/hwchase17/deepagents. Review the README, Code, and Social Media."}]})
+
+for chunk in result:
+ if "agent" in chunk:
+ msg = chunk["agent"]["messages"][0].content.strip()
+ logger.info("=== Agent Message ===\n%s", msg)
+ if "tools" in chunk:
+ call = chunk["tools"]["messages"][0].content.strip()
+ logger.info("+++ Tool Call +++\n%s", call)
+
+logger.info("Finished streaming. Log saved to %s", log_filename)