docs(reference): document NEMOCLAW_OLLAMA_PULL_TIMEOUT env var (#3116)#3122
Conversation
…A#3116) The NEMOCLAW_OLLAMA_PULL_TIMEOUT env var was added by NVIDIA#3037 in src/lib/onboard-ollama-proxy.ts (PULL_TIMEOUT_ENV constant, gate inside getOllamaPullTimeoutMs) but never documented. Users hitting the 30-minute default on slow links cannot discover how to raise it. Add a "Onboard timeouts" subsection under the existing Environment Variables section in docs/reference/commands.md naming the var, default (1800 seconds / 30 minutes), accepted format (integer or float seconds), and the partial-download-resume behavior. Mirror the same addition into the auto-generated skill at .agents/skills/nemoclaw-user-reference/references/commands.md so the skill stays aligned with the source. Closes NVIDIA#3116 Signed-off-by: latenighthackathon <latenighthackathon@users.noreply.github.com>
|
ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Enterprise Run ID: 📒 Files selected for processing (2)
✅ Files skipped from review due to trivial changes (2)
📝 WalkthroughWalkthroughDocumentation for the Environment Variable Documentation
Estimated code review effort🎯 1 (Trivial) | ⏱️ ~3 minutes Poem
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
There was a problem hiding this comment.
🧹 Nitpick comments (1)
docs/reference/commands.md (1)
966-970: ⚡ Quick winSplit multi-sentence lines to one sentence per source line.
Line 966 and Line 970 each contain multiple sentences on a single line; this breaks the docs rule for diff readability.
Suggested edit
-The following environment variables tune onboard-time wall-clock limits. Set them before running `nemoclaw onboard` if a slow connection or large model pull risks tripping the default. +The following environment variables tune onboard-time wall-clock limits. +Set them before running `nemoclaw onboard` if a slow connection or large model pull risks tripping the default. @@ -| `NEMOCLAW_OLLAMA_PULL_TIMEOUT` | `1800` (30 minutes) | Wall-clock timeout for `ollama pull` during onboard, in seconds. Accepts integer or float values. Already-downloaded layers are kept; re-running the pull resumes them. | +| `NEMOCLAW_OLLAMA_PULL_TIMEOUT` | `1800` (30 minutes) | Wall-clock timeout for `ollama pull` during onboard, in seconds. Accepts integer or float values. Already-downloaded layers are kept; re-running the pull resumes them. |As per coding guidelines, “One sentence per line in source (makes diffs readable).”
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@docs/reference/commands.md` around lines 966 - 970, The table rows contain cells with multiple sentences on the same source line (notably the `NEMOCLAW_OLLAMA_PULL_TIMEOUT` row describing the timeout and resume behavior); split those multi-sentence cell contents so each sentence is on its own source line per the "one sentence per line" docs rule—e.g., break the long Purpose cell into separate lines for "Wall-clock timeout for `ollama pull` during onboard, in seconds. Accepts integer or float values." and "Already-downloaded layers are kept; re-running the pull resumes them." to make diffs readable and follow the guideline.
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Nitpick comments:
In `@docs/reference/commands.md`:
- Around line 966-970: The table rows contain cells with multiple sentences on
the same source line (notably the `NEMOCLAW_OLLAMA_PULL_TIMEOUT` row describing
the timeout and resume behavior); split those multi-sentence cell contents so
each sentence is on its own source line per the "one sentence per line" docs
rule—e.g., break the long Purpose cell into separate lines for "Wall-clock
timeout for `ollama pull` during onboard, in seconds. Accepts integer or float
values." and "Already-downloaded layers are kept; re-running the pull resumes
them." to make diffs readable and follow the guideline.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Enterprise
Run ID: 59fedf50-0be8-4c18-8aed-9b4b063d20d4
📒 Files selected for processing (2)
.agents/skills/nemoclaw-user-reference/references/commands.mddocs/reference/commands.md
|
✨ Thanks for submitting this PR that documents the NEMOCLAW_OLLAMA_PULL_TIMEOUT environment variable. This change involves adding a new subsection to the Environment Variables section of the commands reference documentation. Related open PRs: Related open issues: |
There was a problem hiding this comment.
Approving — content is accurate, placement under "Environment Variables" with an "Onboard timeouts" subsection is right, and the table covers default + format + resume behavior cleanly.
Closes #3116.
One small heads-up for next time, not a blocker here: .agents/skills/nemoclaw-user-reference/references/commands.md is autogenerated by scripts/docs-to-skills.py and gets regenerated during release prep. Per docs/CONTRIBUTING.md, contributor PRs should only edit the source under docs/. Your hand-edit happens to match what the generator would produce so it's harmless this time, but in future PRs you can drop it and let the dry-run hook validate it. Thanks!
Summary
#3116 reports that
NEMOCLAW_OLLAMA_PULL_TIMEOUTwas added in code by #3037 (PULL_TIMEOUT_ENVconstant insrc/lib/onboard-ollama-proxy.ts, gate insidegetOllamaPullTimeoutMs) but was never added todocs/reference/commands.md. Users hitting the 30-minute default on slow links cannot discover how to raise it; the only mention is the in-flight error hint when the timeout actually fires.Problem
NEMOCLAW_OLLAMA_PULL_TIMEOUTis read bygetOllamaPullTimeoutMs()(defaults to 30 minutes / 1800 seconds, accepts integer or float seconds, validated for finite-positive). It controls the wall-clock cap onollama pullduring onboard. Without docs, users only learn the variable name from the timeout-error hint after the failure.Changes
docs/reference/commands.md. New table row documents the variable, default (1800/ 30 minutes), accepted format (integer or float seconds), and the partial-download-resume behavior. One-line console example..agents/skills/nemoclaw-user-reference/references/commands.mdso the auto-generated skill stays aligned with the source.Type of Change
Test plan
src/lib/onboard-ollama-proxy.ts:311-330and the error-hint text atpullTimeoutErrorHint.Closes #3116
Signed-off-by: latenighthackathon latenighthackathon@users.noreply.github.com
Summary by CodeRabbit