Repository navigation
chore: sync upstream (docs guide, e2e suite, network_guidance) - #46
Merged
mannuch merged 5 commits intoNov 3, 2026
Merged
Conversation
## Why The README had grown to 600 lines in the order features were added: shell jobs, crash recovery and subscriptions sat under "Testing", agent cleanup notes after the test command, and `rlm.services`, `rlm.hints` and the policy fields were not documented at all. Someone new could not find how the pieces fit, and someone looking for one fact had to read everything. ## What The README becomes a landing page, and the content moves into a guide under `docs/`, where each page opens with what its subject is for. <picture> <source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/PrimeIntellect-ai/nano-rlm/477080dca24ad094c5543bf2dbc1766fd9f00a19/docs/images/overview-dark.svg"> <img alt="An ACP client sends the runtime contract and prompts to rlm. The engine runs the agent loop against the model API and appends to the session ledger; its tool calls run in a persistent IPython kernel, whose rlm.* calls reach the supervisor; the supervisor owns child agents, each with its own engine, kernel and session directory." src="https://raw.githubusercontent.com/PrimeIntellect-ai/nano-rlm/477080dca24ad094c5543bf2dbc1766fd9f00a19/docs/images/overview.svg"> </picture> - **README:** what nano-rlm is, the overview figure above, install, a quickstart ACP client, and an index of the pages. I ran the quickstart against a hosted model with only the endpoint swapped. - **The pages:** | Page | Covers | |---|---| | [Architecture](https://github.com/PrimeIntellect-ai/nano-rlm/blob/477080dca24ad094c5543bf2dbc1766fd9f00a19/docs/architecture.md) | the components, one turn through them, and trust boundaries | | [Runtime contract](https://github.com/PrimeIntellect-ai/nano-rlm/blob/477080dca24ad094c5543bf2dbc1766fd9f00a19/docs/runtime-contract.md) | every runtime field and policy, prompts, steering, snapshots and stop reasons, tracing headers | | [Python API](https://github.com/PrimeIntellect-ai/nano-rlm/blob/477080dca24ad094c5543bf2dbc1766fd9f00a19/docs/python-api.md) | agents, inbox and `wait`, shell jobs, services, watches, history, hints | | [Tools and skills](https://github.com/PrimeIntellect-ai/nano-rlm/blob/477080dca24ad094c5543bf2dbc1766fd9f00a19/docs/tools.md) | native and built-in Python tools, tool packages, `SKILL.md`, MCP, the kernel's environment | | [Context and history](https://github.com/PrimeIntellect-ai/nano-rlm/blob/477080dca24ad094c5543bf2dbc1766fd9f00a19/docs/context.md) | compaction, tool-output truncation, the ledger | | [Kernel and processes](https://github.com/PrimeIntellect-ai/nano-rlm/blob/477080dca24ad094c5543bf2dbc1766fd9f00a19/docs/processes.md) | timeouts, crash recovery, and which processes outlive a call, a restart or the session | | [Split execution](https://github.com/PrimeIntellect-ai/nano-rlm/blob/477080dca24ad094c5543bf2dbc1766fd9f00a19/docs/split-execution.md) | the workspace worker over stdio or TCP, its cost and responsibilities | - **Checked against the code.** Every fact was checked against the code, which corrected a few the README had wrong: - a compaction gives up after `max_compaction_attempts` (5) attempts, not three; - `max_tokens` is a per-agent completion budget; - the native `wait` tool is offered whenever `ipython` is; - the built-in Python tools include `bash`. - **AGENTS.md** asks a change to user-facing behavior to update its page in the same PR. Links, anchors and images were checked across all pages, and the pages render with GitHub's markdown renderer, figures in light and dark. PrimeIntellect-ai#246 added a README paragraph on ACP stop reasons; it is now a table in the runtime contract, keyed by the snapshot's `last_stop_reason`. PrimeIntellect-ai#244 also edits the README's testing section. Whichever merges second moves the end-to-end notes to the new layout.
…-ai#244) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Upstream turned the README into a landing page with a docs/ guide. The fork's README sections move into that layout: the continual harness and prompt optimization get pages of their own, and the fork's changes to compaction, the ledger, the runtime contract, the Python API, tools and split execution go into the pages that now describe them. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…idance (PrimeIntellect-ai#240) Adds `ExecutionPolicy.network_guidance` (default `true`). It controls the two system-prompt sentences that assume outbound network access: - the `!uv pip install <pkg>` hint in the kernel packages line; - the "online solutions or hints specific to this task" clause of the git history guard. A box with no network can set it to `false`, so the prompt doesn't suggest actions that can't work there. The default leaves the prompt byte-identical. Rendered diff with `network_guidance=false`: ```diff -Pre-installed in the kernel venv: requests, …, pydantic. Install extra packages with `!uv pip install <pkg>` in a code cell — that targets the kernel venv (a uv-managed venv with no pip module). +Pre-installed in the kernel venv: requests, …, pydantic. -Do not cheat by using online solutions or hints specific to this task, or by copying or inferring solutions from other branches, … +Do not cheat by copying or inferring solutions from other branches, … ``` The flag is wired like `allow_git`: through the engine, the ACP runtime contract and the session limits snapshot. Sub-agents inherit it with the policy. Tested: 380 passed. The one failure, `tests/test_skills.py::test_valid_bash_skill_metrics`, also fails on main. ruff check and format are clean. 🤖 Generated with [Claude Code](https://claude.com/claude-code) <!-- CURSOR_SUMMARY --> --- > [!NOTE] > **Low Risk** > Default-on policy keeps existing behavior; changes only affect optional prompt strings and runtime metadata fields. > > **Overview** > Adds **`ExecutionPolicy.network_guidance`** (default **`true`**) so air-gapped or no-network runs can omit system-prompt lines that assume outbound access. > > When **`network_guidance`** is **`false`**, **`build_system_prompt`** swaps the kernel packages blurb for a pre-installed-only line (drops **`!uv pip install`**) and uses an offline git history guard that keeps local-history rules but drops the “online solutions” clause. With the default **`true`**, composed prompts stay unchanged. > > The flag is threaded like **`allow_git`**: stored on the engine, passed into prompt construction, exposed on the ACP **`limits`** snapshot and **`_LimitsSnapshot`**, and inherited with the rest of **`ExecutionPolicy`** for sub-agents. Tests cover config/ACP fixtures and a dedicated prompt test for online vs offline wording. > > <sup>Reviewed by [Cursor Bugbot](https://cursor.com/bugbot) for commit ef195f8. Bugbot is set up for automated code reviews on this repo. Configure [here](https://www.cursor.com/dashboard/bugbot).</sup> <!-- /CURSOR_SUMMARY --> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Brings in the network_guidance policy (PrimeIntellect-ai#240). Conflicts were additive: the limits snapshot, engine init and test fixtures keep the fork's compaction and harness fields alongside network_guidance, and the kernel packages line keeps the fork's `--python {sys.executable}` install hint on upstream's split KERNEL_PREINSTALLED_PROMPT prefix. docs/runtime-contract.md gains a network_guidance policy row. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
mannuch
changed the title
mannuch
marked this pull request as ready for review
November 3, 2026 00:16
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
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.
Stacked on #45. Merges three upstream commits:
docs: a short README and a docs/ guide (#245). The README becomes a landing page, and its content moves intodocs/.tests: run a hosted model end to end, local and split (#244). Adds an e2e suite undertests/e2e, run only with-m e2e, and its workflow.feat: add network_guidance policy to drop network-dependent prompt guidance (#240). A policy flag that drops the prompt's network-dependent sentences (the package-install hint and the online-solutions clause of the Git guard).Conflict resolution
Docs guide (PrimeIntellect-ai#245)
README.md was the only conflict. Upstream rewrote the old README content into new pages instead of moving it verbatim, so the fork's README sections are rewritten into that layout:
docs/continual-harness.md: the fork's whole "Continual harness" section, with headings promoted and links pointed at the new pages, and the global-store layout from the old "Session directory" section.docs/prompt-optimization.md: the fork's GEPA section.context.md: the tiered compaction subsections (windows vs branches, rollups, staircase, verbatim tail, pinned prompt, drill-down, budgets);compaction/rolluprecords and theharnesswindow reason in the ledger;hist.blocks/hist.expandin the history example;harness/in the session tree. The checkpoint-prompt sentence now describes the fork's prompt.runtime-contract.md: theharnessandprompt_overridesfields; the three compaction policy fields; therefine-v1prompt meta; therefinements-v1snapshot and harness counts; therefinedstop reason; thecompaction_attempt,refinement_attemptandrefinementedges; and therlm harnessoperator commands.python-api.md:rlm.harness/rlm.refinerows, thehistnaming, and the split-mode episode fallback forhistory().tools.md: the fork's kernel-environment description (inherit minus a blocklist, which is whatbuild_kernel_envdoes), theuv pip install --python {sys.executable}note, and a pointer to authored skills.split-execution.md: the skills mirror replaces "the workspace needs the same files" forSKILL.mdskills.scripts/gepa/RUNBOOK.mdandscripts/refine_eval/RUNBOOK.mdnow point at the new pages instead of README sections.Formatting-only differences from the old README (table alignment, JSON spacing) are dropped in favor of upstream's text.
docs/architecture.mdstill says the kernel environment is rebuilt from an allowlist. That doesn't matchbuild_kernel_env, but the page is unchanged upstream text, so it is left as is.network_guidance (PrimeIntellect-ai#240)
The conflicts in
acp.py,engine.py,prompt.pyandtests/test_acp.pywere additive: the limits snapshot, engine init and test fixtures keep the fork's compaction and harness fields, withnetwork_guidanceplaced right afterallow_gitas upstream has it. The kernel packages line keeps the fork'suv pip install --python {sys.executable}hint on top of upstream's newKERNEL_PREINSTALLED_PROMPTprefix, so withnetwork_guidance=falseit drops to the same preinstalled-only line as upstream. Upstream didn't document the flag, sodocs/runtime-contract.mdgets anetwork_guidancepolicy row.Test plan
uv run pytest tests/: 500 passed. The upstream Linux-onlytest_kernel_recovery.py::test_remote_execution_preserves_state_and_brokers_childrenis deselected, along with the e2e suite (opt-in, needs a hosted model).🤖 Generated with Claude Code