fix(code): make Auto mode generally available - #4957
Merged
Conversation
Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com>
github-actions
Bot
added
dcode
| auto_mode_eligible: Whether Auto mode can be enabled in this session. | ||
| When `False` (e.g. the experimental opt-in is off), the "Enable | ||
| Auto for this thread" option is not offered. | ||
| When `False` (e.g. a sandbox is active), the "Enable Auto for this |
Contributor
Author
There was a problem hiding this comment.
Flagging Mason Daugherty (@mdrxy) - why would this be false when a sandbox is active? Surely there are actions we potentially want to deny even when in a sandbox?
Another question would be - do we really need this auto_mode_eligible attr once we're in GA?
Contributor
Author
There was a problem hiding this comment.
Ok looks like this is a deliberate decision - would you provide some info so I understand?
There was a problem hiding this comment.
Auto is local-only for now on purpose.
Under the hood Auto does more than “skip the approve prompt.” It makes judgment calls using context about this machine’s project: where the repo root is, what the git remote looks like, and whether a path is inside that trusted tree. Those checks are built for a normal local worktree on the host.
A sandbox breaks that assumption. Tools aren’t writing to the same FS the host process is looking at — paths, symlinks, and “where is the project root?” can mean something different inside Daytona/Runloop/etc. If we naively reused the host-side checks, Auto could allow or deny for the wrong reasons. Same issue for the temporary files Auto has the agent use for scratch: those are allocated in the host temp dir today, which isn’t the same place sandbox tools run.
Important distinction: sandbox sessions still get normal Manual approvals in the TUI. We’re only holding back the classifier auto-approve path until it’s honest there.
Making Auto work with sandboxes later isn’t a new product mode. Same Shift+Tab / fallback UI. We’d just need the policy to reason about the sandbox's workspace (remote root, remote path semantics, scratch that actually lives where tools run — or host-temp disabled under sandbox). Once that binding is solid, Auto + sandbox can ship together.
| interactive=True, | ||
| auto_approve=False, | ||
| auto_mode_enabled=False, | ||
| auto_mode_enabled=True, |
Contributor
Author
There was a problem hiding this comment.
There was a problem hiding this comment.
Yep — intentional.
With the experimental gate gone, interactive + no sandbox enables Auto at graph build:
auto_mode_enabled = config.interactive and sandbox_backend is None
So this expectation flips False → True for the default local interactive path. Sandbox (or non-interactive) still keeps it off.
Mason Daugherty (mdrxy)
marked this pull request as ready for review
July 22, 2026 20:34
![open-swe[bot]](https://avatars.githubusercontent.com/in/1395372?s=60&v=4){kind=small}
Remove beta framing from Auto mode help/copy now that the feature is GA, delete needless experimental-flag regression tests, and stop patching the experimental env in interactive --auto-approve coverage.
Mason Daugherty (mdrxy)
changed the title
Auto is now generally available, so the sandbox-unavailable toast no longer misattributes the block to "the local TUI", and residual "beta" labels are dropped from the enabled warnings. Adds coverage for the branches the GA change restructured: Auto wires its middleware when interactive and sandbox-free, is refused with a sandbox or when non-interactive, resolves at startup with EXPERIMENTAL explicitly off, and downgrades to Manual (with a reason) under --sandbox.
github-actions
Bot
added
fix
Mason Daugherty (mdrxy)
pushed a commit
that referenced
this pull request
Jul 24, 2026
> [!CAUTION] > Merging this PR will automatically publish to **PyPI** and create a **GitHub release**. For the full release process, see [`.github/RELEASING.md`](https://github.com/langchain-ai/deepagents/blob/main/.github/RELEASING.md). --- _Release notes preview: keep this section in sync with the package `CHANGELOG.md`. The published GitHub release body is extracted from the merged `CHANGELOG.md` by `release.yml`, not from this PR description._ --- ## [0.1.46](deepagents-code==0.1.45...deepagents-code==0.1.46) (2026-07-24) ### Highlights - Auto mode is now generally available. [#4957](#4957) - Added configurable Auto goal-criteria acceptance. [#4940](#4940) - Improved Auto behavior by authorizing actions from active goal/rubric directives, avoiding redundant approval prompts, showing the enable notice only on first global enable, deduplicating classifier-unavailable transcript spam, logging underlying classifier failures, and reporting classifier timeout budgets. [#5017](#5017) [#4993](#4993) [#5012](#5012) [#5013](#5013) [#5011](#5011) [#5025](#5025) - Added Hooks v2 capability snapshots and session transcripts, and hardened Hooks v2 command execution. [#4916](#4916) [#4918](#4918) [#4917](#4917) - Raised the agent recursion limit to 2000 and made it configurable. [#4994](#4994) ### Improvements and fixes - Let the rubric grader inspect working-directory files, show rubric grader defaults, and improved `/rubric` help and empty-state messaging. [#4835](#4835) [#4966](#4966) [#5015](#5015) - Unified goal activation signaling. [#4980](#4980) - Made Version, Model, and CWD copyable in the Debug Console. [#4975](#4975) - Improved `config get` output when a key is missing. [#4976](#4976) - Aborted YOLO launch on `Ctrl+C`/`Ctrl+D` and made the YOLO warning friendlier for new users. [#4953](#4953) [#4950](#4950) - Updated LangSmith handling: secret redaction is disabled by default, `/trace` now flags empty env overrides that shadow the LangSmith key, and the default US endpoint is no longer treated as a custom target. [#4970](#4970) [#4996](#4996) [#5022](#5022) - Injected OpenAI `prompt_cache_key` for any OpenAI-provider endpoint. [#4995](#4995) - Improved tool and schema presentation: finished calls stay on the live tool-group line, first-party tool schemas now include field descriptions, and `web_search`/`fetch_url` tool descriptions were trimmed. [#4927](#4927) [#5019](#5019) [#5016](#5016) - Omitted `plugins/` and `conversation_history/` from the `/agent` picker. [#4991](#4991) - Made selector modal backdrop dimming consistent. [#4990](#4990) - Restored the `"Server log preserved at:"` notice on exit. [#4999](#4999) - Used the SDK pin as the effective editable version. [#4949](#4949) _End release notes preview._ --- > [!NOTE] > A **New Contributors** section is appended to the GitHub release notes automatically at publish time (see [Release Pipeline](https://github.com/langchain-ai/deepagents/blob/main/.github/RELEASING.md#release-pipeline), step 2). --------- Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Co-authored-by: langchain-oss-automated-triage[bot] <248757908+langchain-oss-automated-triage[bot]@users.noreply.github.com>
Marcelo5444
pushed a commit
to Marcelo5444/deepagents
that referenced
this pull request
Jul 30, 2026
Auto approval mode is now available in the local interactive TUI without setting `DEEPAGENTS_CODE_EXPERIMENTAL`; sandbox and headless sessions still fall back to Manual. --- This removes the experimental environment gate across CLI startup, app eligibility, server graph construction, and agent middleware while preserving the local-only safety constraints. Verified with focused unit tests plus package lint, type, formatting, and import checks. Made by [Open SWE](https://openswe.vercel.app/agents/e0c9c9a9-85ca-7dc8-d759-d89431bc5276) --------- Co-authored-by: Alexander Olsen <13665641+aolsenjazz@users.noreply.github.com> Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> Co-authored-by: Mason Daugherty <mason@langchain.dev> Co-authored-by: Mason Daugherty <github@mdrxy.com>
Marcelo5444
pushed a commit
to Marcelo5444/deepagents
that referenced
this pull request
Jul 30, 2026
> [!CAUTION] > Merging this PR will automatically publish to **PyPI** and create a **GitHub release**. For the full release process, see [`.github/RELEASING.md`](https://github.com/langchain-ai/deepagents/blob/main/.github/RELEASING.md). --- _Release notes preview: keep this section in sync with the package `CHANGELOG.md`. The published GitHub release body is extracted from the merged `CHANGELOG.md` by `release.yml`, not from this PR description._ --- ## [0.1.46](langchain-ai/deepagents@deepagents-code==0.1.45...deepagents-code==0.1.46) (2026-07-24) ### Highlights - Auto mode is now generally available. [langchain-ai#4957](langchain-ai#4957) - Added configurable Auto goal-criteria acceptance. [langchain-ai#4940](langchain-ai#4940) - Improved Auto behavior by authorizing actions from active goal/rubric directives, avoiding redundant approval prompts, showing the enable notice only on first global enable, deduplicating classifier-unavailable transcript spam, logging underlying classifier failures, and reporting classifier timeout budgets. [langchain-ai#5017](langchain-ai#5017) [langchain-ai#4993](langchain-ai#4993) [langchain-ai#5012](langchain-ai#5012) [langchain-ai#5013](langchain-ai#5013) [langchain-ai#5011](langchain-ai#5011) [langchain-ai#5025](langchain-ai#5025) - Added Hooks v2 capability snapshots and session transcripts, and hardened Hooks v2 command execution. [langchain-ai#4916](langchain-ai#4916) [langchain-ai#4918](langchain-ai#4918) [langchain-ai#4917](langchain-ai#4917) - Raised the agent recursion limit to 2000 and made it configurable. [langchain-ai#4994](langchain-ai#4994) ### Improvements and fixes - Let the rubric grader inspect working-directory files, show rubric grader defaults, and improved `/rubric` help and empty-state messaging. [langchain-ai#4835](langchain-ai#4835) [langchain-ai#4966](langchain-ai#4966) [langchain-ai#5015](langchain-ai#5015) - Unified goal activation signaling. [langchain-ai#4980](langchain-ai#4980) - Made Version, Model, and CWD copyable in the Debug Console. [langchain-ai#4975](langchain-ai#4975) - Improved `config get` output when a key is missing. [langchain-ai#4976](langchain-ai#4976) - Aborted YOLO launch on `Ctrl+C`/`Ctrl+D` and made the YOLO warning friendlier for new users. [langchain-ai#4953](langchain-ai#4953) [langchain-ai#4950](langchain-ai#4950) - Updated LangSmith handling: secret redaction is disabled by default, `/trace` now flags empty env overrides that shadow the LangSmith key, and the default US endpoint is no longer treated as a custom target. [langchain-ai#4970](langchain-ai#4970) [langchain-ai#4996](langchain-ai#4996) [langchain-ai#5022](langchain-ai#5022) - Injected OpenAI `prompt_cache_key` for any OpenAI-provider endpoint. [langchain-ai#4995](langchain-ai#4995) - Improved tool and schema presentation: finished calls stay on the live tool-group line, first-party tool schemas now include field descriptions, and `web_search`/`fetch_url` tool descriptions were trimmed. [langchain-ai#4927](langchain-ai#4927) [langchain-ai#5019](langchain-ai#5019) [langchain-ai#5016](langchain-ai#5016) - Omitted `plugins/` and `conversation_history/` from the `/agent` picker. [langchain-ai#4991](langchain-ai#4991) - Made selector modal backdrop dimming consistent. [langchain-ai#4990](langchain-ai#4990) - Restored the `"Server log preserved at:"` notice on exit. [langchain-ai#4999](langchain-ai#4999) - Used the SDK pin as the effective editable version. [langchain-ai#4949](langchain-ai#4949) _End release notes preview._ --- > [!NOTE] > A **New Contributors** section is appended to the GitHub release notes automatically at publish time (see [Release Pipeline](https://github.com/langchain-ai/deepagents/blob/main/.github/RELEASING.md#release-pipeline), step 2). --------- Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Co-authored-by: langchain-oss-automated-triage[bot] <248757908+langchain-oss-automated-triage[bot]@users.noreply.github.com>
23 tasks
Prekshi Vyas (prekshivyas)
added a commit
to NVIDIA/NemoClaw
that referenced
this pull request
Aug 19, 2026
<!-- patch-walker:action=sha256:4664592e4dabe18250a1e41ea73d7121d7ef19dd4cbc6c01bfe297c519a75974 --> <!-- patch-walker:manifest=sha256:95728f38d033eae61d4905337245c21043f166f5ef57eaac4908f89e2cc24d7f --> <!-- patch-walker:dependency=LangChain Deep Agents Code --> <!-- patch-walker:target=0.1.55 --> <!-- markdownlint-disable MD041 --> ## Summary Updates LangChain Deep Agents Code from 0.1.34 to 0.1.55 using the sealed NemoPin migration evidence. The implementation and requested-review fixes are complete; the PR is ready for maintainer re-review. <!-- 1-3 plain sentences: what changes and why. Describe before-and-after behavior when it applies. Follow the NemoClaw Writing Guide: https://github.com/NVIDIA/NemoClaw/blob/main/WRITING.md. Do not add unrelated prose cleanup. --> ## Related Issue No issue closure is claimed. <!-- Fixes #NNN or Closes #NNN. Remove this section if none. --> ## Changes\n\n- Keeps the dependency migration scoped to LangChain Deep Agents Code and its onboarding, validation, documentation, and managed-image checks.\n- Migrates the exact base `5bb69ed66947fbd2fab6133748866567eb520692` across 21 adjacent release ranges.\n- Changed paths: `.github/workflows/managed-images.yaml`, `agents/langchain-deepagents-code/Dockerfile`, `agents/langchain-deepagents-code/Dockerfile.base`, `agents/langchain-deepagents-code/dcode-wrapper.sh`, `agents/langchain-deepagents-code/dependency-review.md`, `agents/langchain-deepagents-code/manifest.yaml`, `agents/langchain-deepagents-code/patch-managed-deepagents-code.py`, `agents/langchain-deepagents-code/profile-plugin/pyproject.toml`, `agents/langchain-deepagents-code/profile-plugin/src/nemoclaw_deepagents_profile/__init__.py`, `agents/langchain-deepagents-code/progressive_tool_disclosure.py`, `agents/langchain-deepagents-code/requirements.in`, `agents/langchain-deepagents-code/requirements.lock`, `agents/langchain-deepagents-code/start.sh`, `agents/langchain-deepagents-code/validate-nemotron-ultra-profile.py`, `agents/langchain-deepagents-code/validate-observability.py`, `agents/langchain-deepagents-code/validate-progressive-tool-disclosure.py`, `docs/deployment/set-up-mcp-bridge.mdx`, `docs/get-started/quickstart-langchain-deepagents-code.mdx`, `src/lib/actions/sandbox/rebuild-flow-helpers.test.ts`, `src/lib/agent/base-image.test.ts`, `src/lib/agent/deep-agents-code-base-image.test.ts`, `src/lib/agent/onboard-terminal-fixtures.test.ts`, `src/lib/agent/onboard-terminal-fixtures.ts`, `src/lib/agent/onboard-terminal.test.ts`, `src/lib/inference/onboard-probes.test.ts`, `src/lib/inference/onboard-probes.ts`, `src/lib/inference/openai-validation-session.ts`, `src/lib/onboard.ts`, `src/lib/onboard/created-sandbox-finalization.test.ts`, `src/lib/onboard/created-sandbox-finalization.ts`, `src/lib/onboard/dcode-selection-drift.test.ts`, `src/lib/onboard/dcode-selection-drift.ts`, `src/lib/onboard/inference-selection-validation.test.ts`, `src/lib/onboard/machine/handlers/sandbox-checkpoint-crash-recovery.test.ts`, `src/lib/onboard/machine/handlers/sandbox.ts`, `src/lib/onboard/sandbox-lifecycle.test.ts`, `src/lib/onboard/sandbox-lifecycle.ts`, `src/lib/sandbox-base-image-agent-resolution.test.ts`, `src/lib/sandbox-base-image-release-resolution.test.ts`, `src/lib/sandbox-base-image/resolution-key.test.ts`, `test/Dockerfile.dcode-profile-missing-dependencies`, `test/cli/connect-terminal-agent.test.ts`, `test/deepagents-code-tui-startup-check.test.ts`, `test/e2e/e2e-cloud-experimental/checks/03-deepagents-code-nemotron-ultra-profile.sh`, `test/e2e/e2e-cloud-experimental/checks/04-deepagents-code-fresh-reonboard.sh`, `test/e2e/e2e-cloud-experimental/checks/10-deepagents-code-tui-startup.sh`, `test/e2e/e2e-cloud-experimental/checks/12-deepagents-code-thread-auto-approval.sh`, `test/e2e/lib/select-authorized-chat-model.mts`, `test/e2e/live/mcp-bridge-servers.ts`, `test/e2e/support/authorized-chat-model-selection.test.ts`, `test/fixtures/deepagents-progressive-disclosure-harness.py`, `test/fixtures/langchain-deepagents-code/server.py`, `test/helpers/langchain-deepagents-code-patch-fixture.ts`, `test/helpers/managed-image-buildless-e2e.ts`, `test/issue-5667-hosted-inference-model-namespace.test.ts`, `test/langchain-deepagents-code-direct-module-patch.test.ts`, `test/langchain-deepagents-code-image.test.ts`, `test/langchain-deepagents-code-nemotron-profile-plugin.test.ts`, `test/langchain-deepagents-code-progressive-tool-disclosure.test.ts`, `test/managed-image-publication-workflow.test.ts`, `test/managed-image-staging-qa-workflow.test.ts`, `test/mcp-bridge-servers.test.ts`, `test/onboard-mcp-observability-redirect.test.ts`, `test/onboard-prepared-build-context.test.ts`, `test/onboard-terminal-dashboard.test.ts`\n\n### Release ranges | Range | Commits | State | Concerns | |---|---|---|---| | 0.1.34 → 0.1.35 | `bd9bafaad3f5` → `09daab5772ff` | published | 1 | | 0.1.35 → 0.1.36 | `09daab5772ff` → `2f56309d821d` | published | 1 | | 0.1.36 → 0.1.37 | `2f56309d821d` → `def6369aed1a` | published | 2 | | 0.1.37 → 0.1.38 | `def6369aed1a` → `4338671aa1d9` | published | 4 | | 0.1.38 → 0.1.39 | `4338671aa1d9` → `8eb909a59b82` | published | 1 | | 0.1.39 → 0.1.40 | `8eb909a59b82` → `019489edb9c0` | published | 6 | | 0.1.40 → 0.1.41 | `019489edb9c0` → `d46a2cb033b8` | published | 4 | | 0.1.41 → 0.1.42 | `d46a2cb033b8` → `18679a1a88a3` | published | 3 | | 0.1.42 → 0.1.43 | `18679a1a88a3` → `e14e0adcbe78` | published | 2 | | 0.1.43 → 0.1.44 | `e14e0adcbe78` → `2b9cd08f0492` | published | 5 | | 0.1.44 → 0.1.45 | `2b9cd08f0492` → `7794b61a6e76` | published | 4 | | 0.1.45 → 0.1.46 | `7794b61a6e76` → `efa86c51fedd` | published | 1 | | 0.1.46 → 0.1.47 | `efa86c51fedd` → `8aa29ddc2833` | published | 3 | | 0.1.47 → 0.1.48 | `8aa29ddc2833` → `803b8329db7d` | published | 3 | | 0.1.48 → 0.1.49 | `803b8329db7d` → `44910bc2ef3f` | published | 0 | | 0.1.49 → 0.1.50 | `44910bc2ef3f` → `63adb9645687` | published | 2 | | 0.1.50 → 0.1.51 | `63adb9645687` → `d2b663fca277` | published | 0 | | 0.1.51 → 0.1.52 | `d2b663fca277` → `b428644d31dd` | published | 3 | | 0.1.52 → 0.1.53 | `b428644d31dd` → `0bd15dc0e1c5` | published | 2 | | 0.1.53 → 0.1.54 | `0bd15dc0e1c5` → `81258067f4c7` | published | 0 | | 0.1.54 → 0.1.55 | `81258067f4c7` → `80fe3d3cbcd2` | published | 9 | ### Concern dispositions | Concern | Surface | Planned disposition | Failure prevented | Remaining gate | |---|---|---|---|---| | `langchain-deep-agents-code-0.1.34..0.1.35-lifecycle-state-1` | lifecycle state | test | 0.1.55 reports lifecycle state: ### Features - Added a `/context` usage report for inspecting context consumption ([#5407](langchain-ai/deepagents#5407... | none | | `langchain-deep-agents-code-0.1.35..0.1.36-lifecycle-state-1` | lifecycle state | test | 0.1.54 reports lifecycle state: ### Features - Added Meta `muse-spark-1.2` to the model switcher ([#5389](langchain-ai/deepagents#5389)). - Improved di... | none | | `langchain-deep-agents-code-0.1.36..0.1.37-lifecycle-state-1` | lifecycle state | test | 0.1.53 reports lifecycle state: ### Features - Added pricing coverage with Baseten built-in overrides and local fallback overrides when `genai-prices` is missing data ([#5312](h... | none | | `langchain-deep-agents-code-0.1.36..0.1.37-runtime-topology-2` | runtime topology | test | 0.1.53 reports runtime topology: ### Features - Added pricing coverage with Baseten built-in overrides and local fallback overrides when `genai-prices` is missing data ([#5312](... | none | | `langchain-deep-agents-code-0.1.37..0.1.38-compatibility-change-1` | compatibility change | test | 0.1.52 reports compatibility change: ### Features - Hooks v2 is now generally available, with support for loading hooks from installed plugins. ([#5307](https://github.com/langc... | none | | `langchain-deep-agents-code-0.1.37..0.1.38-configuration-2` | configuration | test | 0.1.52 reports configuration: ### Features - Hooks v2 is now generally available, with support for loading hooks from installed plugins. ([#5307](https://github.com/langchain-ai... | none | | `langchain-deep-agents-code-0.1.37..0.1.38-execution-control-3` | execution control | guard | 0.1.52 reports execution control: ### Features - Hooks v2 is now generally available, with support for loading hooks from installed plugins. ([#5307](https://github.com/langchai... | none | | `langchain-deep-agents-code-0.1.37..0.1.38-lifecycle-state-4` | lifecycle state | test | 0.1.52 reports lifecycle state: ### Features - Hooks v2 is now generally available, with support for loading hooks from installed plugins. ([#5307](https://github.com/langchain-... | none | | `langchain-deep-agents-code-0.1.38..0.1.39-configuration-1` | configuration | test | 0.1.51 reports configuration: ### Features - The status bar and usage view now show the running session cost. ([#5036](langchain-ai/deepagents#5036)) -... | none | | `langchain-deep-agents-code-0.1.39..0.1.40-compatibility-change-1` | compatibility change | test | 0.1.50 reports compatibility change: ### Highlights - Added project hooks workspace trust and expanded Hooks v2 support with client and server lifecycle events plus runtime feed... | none | | `langchain-deep-agents-code-0.1.39..0.1.40-execution-control-2` | execution control | guard | 0.1.50 reports execution control: ### Highlights - Added project hooks workspace trust and expanded Hooks v2 support with client and server lifecycle events plus runtime feedbac... | none | | `langchain-deep-agents-code-0.1.39..0.1.40-lifecycle-state-3` | lifecycle state | test | 0.1.50 reports lifecycle state: ### Highlights - Added project hooks workspace trust and expanded Hooks v2 support with client and server lifecycle events plus runtime feedback ... | none | | `langchain-deep-agents-code-0.1.39..0.1.40-packaging-artifact-4` | packaging artifact | test | 0.1.50 reports packaging artifact: ### Highlights - Added project hooks workspace trust and expanded Hooks v2 support with client and server lifecycle events plus runtime feedba... | none | | `langchain-deep-agents-code-0.1.39..0.1.40-runtime-topology-5` | runtime topology | test | 0.1.50 reports runtime topology: ### Highlights - Added project hooks workspace trust and expanded Hooks v2 support with client and server lifecycle events plus runtime feedback... | none | | `langchain-deep-agents-code-0.1.39..0.1.40-security-identity-6` | security identity | guard | 0.1.50 reports security identity: ### Highlights - Added project hooks workspace trust and expanded Hooks v2 support with client and server lifecycle events plus runtime feedbac... | none | | `langchain-deep-agents-code-0.1.40..0.1.41-compatibility-change-1` | compatibility change | test | 0.1.49 reports compatibility change: ### Features - Added recognition for LangSmith Gateway credentials. ([#5042](langchain-ai/deepagents#5042)) - Adde... | none | | `langchain-deep-agents-code-0.1.40..0.1.41-execution-control-2` | execution control | guard | 0.1.49 reports execution control: ### Features - Added recognition for LangSmith Gateway credentials. ([#5042](langchain-ai/deepagents#5042)) - Added s... | none | | `langchain-deep-agents-code-0.1.40..0.1.41-lifecycle-state-3` | lifecycle state | test | 0.1.49 reports lifecycle state: ### Features - Added recognition for LangSmith Gateway credentials. ([#5042](langchain-ai/deepagents#5042)) - Added sla... | none | | `langchain-deep-agents-code-0.1.40..0.1.41-runtime-topology-4` | runtime topology | test | 0.1.49 reports runtime topology: ### Features - Added recognition for LangSmith Gateway credentials. ([#5042](langchain-ai/deepagents#5042)) - Added sl... | none | | `langchain-deep-agents-code-0.1.41..0.1.42-compatibility-change-1` | compatibility change | test | 0.1.48 reports compatibility change: ### Features - Added Fireworks `kimi-k3`, GLM-5.2-Fast, and Kimi-K3 to model selection and recommended models. ([#5082](https://github.com/l... | none | | `langchain-deep-agents-code-0.1.41..0.1.42-execution-control-2` | execution control | guard | 0.1.48 reports execution control: ### Features - Added Fireworks `kimi-k3`, GLM-5.2-Fast, and Kimi-K3 to model selection and recommended models. ([#5082](https://github.com/lang... | none | | `langchain-deep-agents-code-0.1.41..0.1.42-lifecycle-state-3` | lifecycle state | test | 0.1.48 reports lifecycle state: ### Features - Added Fireworks `kimi-k3`, GLM-5.2-Fast, and Kimi-K3 to model selection and recommended models. ([#5082](https://github.com/langch... | none | | `langchain-deep-agents-code-0.1.42..0.1.43-compatibility-change-1` | compatibility change | test | 0.1.47 reports compatibility change: ### Features - Added `yolo` mode to the `Shift+Tab` approval cycle ([#5035](langchain-ai/deepagents#5035)). - Show... | none | | `langchain-deep-agents-code-0.1.42..0.1.43-execution-control-2` | execution control | guard | 0.1.47 reports execution control: ### Features - Added `yolo` mode to the `Shift+Tab` approval cycle ([#5035](langchain-ai/deepagents#5035)). - Show th... | none | | `langchain-deep-agents-code-0.1.43..0.1.44-compatibility-change-1` | compatibility change | test | 0.1.46 reports compatibility change: ### Highlights - Auto mode is now generally available. [#4957](langchain-ai/deepagents#4957) - Added configurable ... | none | | `langchain-deep-agents-code-0.1.43..0.1.44-configuration-2` | configuration | test | 0.1.46 reports configuration: ### Highlights - Auto mode is now generally available. [#4957](langchain-ai/deepagents#4957) - Added configurable Auto go... | none | | `langchain-deep-agents-code-0.1.43..0.1.44-execution-control-3` | execution control | guard | 0.1.46 reports execution control: ### Highlights - Auto mode is now generally available. [#4957](langchain-ai/deepagents#4957) - Added configurable Aut... | none | | `langchain-deep-agents-code-0.1.43..0.1.44-protocol-schema-4` | protocol schema | test | 0.1.46 reports protocol schema: ### Highlights - Auto mode is now generally available. [#4957](langchain-ai/deepagents#4957) - Added configurable Auto ... | none | | `langchain-deep-agents-code-0.1.43..0.1.44-security-identity-5` | security identity | guard | 0.1.46 reports security identity: ### Highlights - Auto mode is now generally available. [#4957](langchain-ai/deepagents#4957) - Added configurable Aut... | none | | `langchain-deep-agents-code-0.1.44..0.1.45-compatibility-change-1` | compatibility change | test | 0.1.45 reports compatibility change: ### Features - Added the Hooks v2 execution engine and typed hooks data models ([#4880](langchain-ai/deepagents#48... | none | | `langchain-deep-agents-code-0.1.44..0.1.45-execution-control-2` | execution control | guard | 0.1.45 reports execution control: ### Features - Added the Hooks v2 execution engine and typed hooks data models ([#4880](langchain-ai/deepagents#4880)... | none | | `langchain-deep-agents-code-0.1.44..0.1.45-lifecycle-state-3` | lifecycle state | test | 0.1.45 reports lifecycle state: ### Features - Added the Hooks v2 execution engine and typed hooks data models ([#4880](langchain-ai/deepagents#4880), ... | none | | `langchain-deep-agents-code-0.1.44..0.1.45-packaging-artifact-4` | packaging artifact | test | 0.1.45 reports packaging artifact: ### Features - Added the Hooks v2 execution engine and typed hooks data models ([#4880](langchain-ai/deepagents#4880... | none | | `langchain-deep-agents-code-0.1.45..0.1.46-runtime-topology-1` | runtime topology | test | 0.1.44 reports runtime topology: ### Bug Fixes - Improved approval handling by hiding the `Auto` option when it isn't eligible and moving Auto mode path checks off the event loo... | none | | `langchain-deep-agents-code-0.1.46..0.1.47-compatibility-change-1` | compatibility change | test | 0.1.43 reports compatibility change: ### Features - Added classifier-backed Auto approval mode behind `DEEPAGENTS_CODE_EXPERIMENTAL=1` ([#4804](https://github.com/langchain-ai/d... | none | | `langchain-deep-agents-code-0.1.46..0.1.47-execution-control-2` | execution control | guard | 0.1.43 reports execution control: ### Features - Added classifier-backed Auto approval mode behind `DEEPAGENTS_CODE_EXPERIMENTAL=1` ([#4804](https://github.com/langchain-ai/deep... | none | | `langchain-deep-agents-code-0.1.46..0.1.47-lifecycle-state-3` | lifecycle state | test | 0.1.43 reports lifecycle state: ### Features - Added classifier-backed Auto approval mode behind `DEEPAGENTS_CODE_EXPERIMENTAL=1` ([#4804](https://github.com/langchain-ai/deepag... | none | | `langchain-deep-agents-code-0.1.47..0.1.48-compatibility-change-1` | compatibility change | test | 0.1.42 reports compatibility change: ### Features - Plugins are now generally available. ([#4797](langchain-ai/deepagents#4797)) - Added search to the ... | none | | `langchain-deep-agents-code-0.1.47..0.1.48-execution-control-2` | execution control | guard | 0.1.42 reports execution control: ### Features - Plugins are now generally available. ([#4797](langchain-ai/deepagents#4797)) - Added search to the plu... | none | | `langchain-deep-agents-code-0.1.47..0.1.48-lifecycle-state-3` | lifecycle state | test | 0.1.42 reports lifecycle state: ### Features - Plugins are now generally available. ([#4797](langchain-ai/deepagents#4797)) - Added search to the plugi... | none | | `langchain-deep-agents-code-0.1.49..0.1.50-compatibility-change-1` | compatibility change | test | 0.1.40 reports compatibility change: ### Features - Added plugin marketplace support ([#4554](langchain-ai/deepagents#4554)). - Added an “always allow”... | none | | `langchain-deep-agents-code-0.1.49..0.1.50-execution-control-2` | execution control | guard | 0.1.40 reports execution control: ### Features - Added plugin marketplace support ([#4554](langchain-ai/deepagents#4554)). - Added an “always allow” op... | none | | `langchain-deep-agents-code-0.1.51..0.1.52-compatibility-change-1` | compatibility change | test | 0.1.38 reports compatibility change: ### Features * Improve `/goal` criteria UX ([#4694](langchain-ai/deepagents#4694)) ([06f46ff](https://github.com/l... | none | | `langchain-deep-agents-code-0.1.51..0.1.52-configuration-2` | configuration | test | 0.1.38 reports configuration: ### Features * Improve `/goal` criteria UX ([#4694](langchain-ai/deepagents#4694)) ([06f46ff](https://github.com/langchai... | none | | `langchain-deep-agents-code-0.1.51..0.1.52-lifecycle-state-3` | lifecycle state | test | 0.1.38 reports lifecycle state: ### Features * Improve `/goal` criteria UX ([#4694](langchain-ai/deepagents#4694)) ([06f46ff](https://github.com/langch... | none | | `langchain-deep-agents-code-0.1.52..0.1.53-configuration-1` | configuration | test | 0.1.37 reports configuration: ### Features * Add Meta model provider ([#4650](langchain-ai/deepagents#4650)) ([70829c5](https://github.com/langchain-ai... | none | | `langchain-deep-agents-code-0.1.52..0.1.53-security-identity-2` | security identity | guard | 0.1.37 reports security identity: ### Features * Add Meta model provider ([#4650](langchain-ai/deepagents#4650)) ([70829c5](https://github.com/langchai... | none | | `langchain-deep-agents-code-0.1.54..0.1.55-compatibility-change-1` | compatibility change | test | 0.1.35 reports compatibility change: ### Features * Restore interrupted prompt to input on ESC ([#4544](langchain-ai/deepagents#4544)) ([fccf037](https... | none | | `langchain-deep-agents-code-0.1.54..0.1.55-configuration-2` | configuration | test | 0.1.35 reports configuration: ### Features * Restore interrupted prompt to input on ESC ([#4544](langchain-ai/deepagents#4544)) ([fccf037](https://gith... | none | | `langchain-deep-agents-code-0.1.54..0.1.55-execution-control-3` | execution control | guard | 0.1.35 reports execution control: ### Features * Restore interrupted prompt to input on ESC ([#4544](langchain-ai/deepagents#4544)) ([fccf037](https://... | none | | `langchain-deep-agents-code-0.1.54..0.1.55-lifecycle-state-4` | lifecycle state | test | 0.1.35 reports lifecycle state: ### Features * Restore interrupted prompt to input on ESC ([#4544](langchain-ai/deepagents#4544)) ([fccf037](https://gi... | none | | `langchain-deep-agents-code-0.1.54..0.1.55-protocol-schema-5` | protocol schema | test | 0.1.35 reports protocol schema: ### Features * Restore interrupted prompt to input on ESC ([#4544](langchain-ai/deepagents#4544)) ([fccf037](https://gi... | none | | `langchain-deep-agents-code-0.1.54..0.1.55-code-impact-configuration-1` | mapped configuration | test | The exact-ref diff reports configuration changes in .pre-commit-config.yaml, libs/acp/deepagents_acp/server.py, libs/acp/tests/test_agent.py. Mapped NemoClaw examples: src/lib/a... | none | | `langchain-deep-agents-code-0.1.54..0.1.55-code-impact-contract-or-schema-2` | mapped contract or schema | test | The exact-ref diff reports contract or schema changes in libs/acp/deepagents_acp/_version.py, libs/acp/deepagents_acp/server.py, libs/code/deepagents_code/_ask_user_types.py. Ma... | none | | `langchain-deep-agents-code-0.1.54..0.1.55-code-impact-security-3` | mapped security | test | The exact-ref diff reports security changes in libs/code/deepagents_code/_cli_context.py, libs/code/deepagents_code/_env_vars.py, libs/code/deepagents_code/_repository_bounds.py... | none | | `langchain-deep-agents-code-0.1.54..0.1.55-code-impact-test-4` | mapped test | test | The exact-ref diff reports test changes in libs/acp/tests/chat_model.py, libs/acp/tests/test_agent.py, libs/code/tests/integration_tests/benchmarks/test_local_context_benchmarks... | none | ### Immutable artifacts | Artifact | SHA-256 | |---|---| | deepagents_code-0.1.55-py3-none-any.whl | `3a0d3e332f132d0e…` | | deepagents_code-0.1.55.tar.gz | `91c30b62cb96d5e8…` | ### Validation receipt | Gate | Current result | |---|---| | targeted | Pass — 130 affected local tests, commit hooks, growth guardrails, and pre-push typechecks passed on the unchanged PR patch now at `78268b19e` | | full-e2e | In progress on the latest PR commit — all required checks pass; three non-required managed-image jobs are still running | <!-- List concrete changes. If this adds an abstraction, configuration, fallback, migration, or compatibility path, name its current requirement and consumer, explain why a direct change is insufficient, and identify the test that protects it. --> ## Type of Change - [ ] Code change (feature, bug fix, or refactor) - [x] Code change with doc updates - [ ] Doc only (prose changes, no code sample modifications) - [ ] Doc only (includes code sample changes) ## Quality Gates - [x] Tests added or updated for changed behavior - [ ] Existing tests cover changed behavior — justification: - [ ] Tests not applicable — justification: - [x] Docs updated for user-facing behavior changes - [ ] Docs not applicable — justification: - [x] Sensitive paths changed (security, policy, credentials, preflight, onboarding, inference, runner, sandbox, or messaging) - [x] Sensitive-path review completed or maintainer-approved waiver recorded — CodeRabbit completed successfully on the unchanged PR patch and all inline review threads are resolved; human re-review remains requested. - [ ] Non-success, skipped, or missing CI check accepted by maintainer — check name, approval link, and follow-up issue: ## Documentation Writer Review - [x] Documentation writer subagent reviewed the completed changes - Result: `docs-updated` - Evidence: `docs/deployment/set-up-mcp-bridge.mdx` and `docs/get-started/quickstart-langchain-deepagents-code.mdx` accurately document the changed Deep Agents Code behavior and versions. The documented deepagents-code 0.1.55 and deepagents 0.7.5 versions match the manifest, inputs, lockfile, and profile plugin. `npm run docs` passed with 0 errors and two pre-existing Fern warnings. - Agent: Codex Desktop <!-- docs-review-head-sha: 78268b1 --> <!-- docs-review-agents-blob-sha: e30afb2 --> ## DGX Station Hardware Evidence <!-- Required only when scripts/prepare-dgx-station-host.sh changes. Maintainers must review the linked evidence before approving or merging. This is human-reviewed evidence, not authenticated hardware provenance. Exceptional bypasses use existing repository governance and must be documented on the PR. --> - [ ] Tested on DGX Station - Tested commit: - Station profile/scenario: - Result: - Supporting evidence: ## Verification <!-- Check each applicable item only when supported by the requested evidence. Run targeted tests once per relevant change set and rerun after later edits or hook autofixes that can affect the tested behavior. Do not rerun hook-covered checks. --> - [x] PR description includes a `Signed-off-by:` line and all 81 commits appear as `Verified` in GitHub - [x] Normal `pre-commit`, `commit-msg`, and `pre-push` hooks passed on the unchanged PR patch now at `78268b19e` - [x] Targeted behavior tests pass for the current change set — affected local suites: 130/130; growth guardrails passed - [ ] Applicable broad gate passed — all required GitHub checks pass; non-required managed-image validation is still in progress - [x] Quality Gates section completed with required justifications - [x] No secrets, API keys, or credentials committed; gitleaks passed - [ ] `npm run docs` builds without warnings (doc changes only) - [ ] Doc pages follow the [style guide](https://github.com/NVIDIA/NemoClaw/blob/main/docs/CONTRIBUTING.md) (doc changes only) - [ ] New doc pages include SPDX header and frontmatter (new pages only) --- <!-- DCO sign-off is required in this PR description, and every commit must appear as Verified in GitHub. Run: git config user.name && git config user.email --> Signed-off-by: Prekshi Vyas <prekshiv@nvidia.com> <!-- patch-walker-status:start --> ## NemoPatch latest PR commit status - Status: **ready for maintainer re-review** - Latest PR commit: `78268b19e8229ffbc2a0591ad310caf96aab7fa6` - Checked: 2026-08-19T19:39:34Z - Merge: GitHub reports **MERGEABLE**; the local patch check against current `upstream/main` is clean. - Review: all 8 inline review threads are resolved and no new review finding was posted for the unchanged patch. Human approval remains required. - CI: all required checks pass. Three broader non-required managed-image jobs are still running with no deterministic failure at this check. - External advisor infrastructure: GPT-5.6 Terra and Nemotron 3 Ultra failed with `investigate omitted required analysis`; the trusted publisher reports 0 blockers, 0 warnings, 0 suggestions, and no follow-up needed. <!-- patch-walker-status:end --> <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Upgraded Deep Agents Code support to 0.1.55 and Deep Agents to 0.7.5. * Expanded approval controls with manual, automatic, startup, and YOLO modes. * Tool search and discovery now include registered tools across agents and subagents. * Added gateway-aware sandbox operations and authorized chat-model selection during onboarding. * **Bug Fixes** * Improved sandbox crash recovery and identity detection. * Strengthened MCP result validation and protected private inference endpoints. * **Documentation** * Updated setup and quickstart guidance for supported versions. <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Signed-off-by: Prekshi Vyas <prekshiv@nvidia.com> Signed-off-by: Prekshiv <prekshiv@nvidia.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Auto approval mode is now available in the local interactive TUI without setting
DEEPAGENTS_CODE_EXPERIMENTAL; sandbox and headless sessions still fall back to Manual.This removes the experimental environment gate across CLI startup, app eligibility, server graph construction, and agent middleware while preserving the local-only safety constraints. Verified with focused unit tests plus package lint, type, formatting, and import checks.
Made by Open SWE