docs: refresh v0.0.62 release docs#5157
Merged
Merged
Conversation
Contributor
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Enterprise Run ID: 📒 Files selected for processing (1)
🚧 Files skipped from review as they are similar to previous changes (1)
📝 WalkthroughWalkthroughThis PR adds and updates documentation across NemoClaw's agent skills and main documentation sites to describe v0.0.62 features: local inference endpoint verification on Linux Docker-driver GPU sandboxes, PTY filesystem ( Changesv0.0.62 Release Documentation
Sequence Diagram(s)sequenceDiagram
participant Onboarder
participant SandboxRuntime
participant InferenceEndpoint
Onboarder->>SandboxRuntime: start sandbox
SandboxRuntime->>InferenceEndpoint: GET https://inference.local/v1/models (from inside sandbox)
InferenceEndpoint-->>SandboxRuntime: HTTP 2xx (success) or non-2xx (failure)
SandboxRuntime-->>Onboarder: report endpoint + recovery steps if non-2xx
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~20 minutes Possibly related PRs
Suggested labels
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 |
Contributor
|
🌿 Preview your docs: https://nvidia-preview-pr-5157.docs.buildwithfern.com/nemoclaw |
Contributor
E2E Advisor RecommendationRequired E2E: None Full advisor summaryE2E Recommendation AdvisorBase: Required E2E
Optional E2E
New E2E recommendations
|
Contributor
E2E Scenario Advisor RecommendationRequired scenario E2E: None Full scenario advisor summaryE2E Scenario AdvisorBase: Required scenario E2E
Optional scenario E2E
Relevant changed files
|
Contributor
PR Review AdvisorFindings: 1 needs attention, 3 worth checking, 0 nice ideas Review findings🛠️ Needs attention
🔎 Worth checking
🌱 Nice ideas
Consider writing more tests for
Since last review detailsCurrent findings:
This is an automated advisory review. A human maintainer must make the final merge decision. |
Contributor
There was a problem hiding this comment.
Actionable comments posted: 3
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (2)
.agents/skills/nemoclaw-user-reference/references/network-policies.md (1)
15-24:⚠️ Potential issue | 🟠 Major | ⚡ Quick winAutogenerated skill file should not be edited directly.
Per coding guidelines, files matching
.agents/skills/nemoclaw-user-*/*.mdare autogenerated and must never be edited directly. These changes should originate from the source docs (e.g.,docs/) and be regenerated viascripts/docs-to-skills.py.Verify that the corresponding source documentation has been updated with the
/dev/ptsfilesystem changes, then regenerate the skills rather than manually editing this file.Based on coding guidelines: "Autogenerated user skills under
.agents/skills/nemoclaw-user-*/*.mdmust never be edited directly"🤖 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 @.agents/skills/nemoclaw-user-reference/references/network-policies.md around lines 15 - 24, This autogenerated markdown under the nemoclaw-user skill was edited directly; instead, update the canonical source docs that feed the skill generator (e.g., the docs/ entry that documents the filesystem table and the /dev/pts change), then run the regeneration script scripts/docs-to-skills.py so the change appears in the generated .agents/skills/nemoclaw-user-*/*.md files; do not commit manual edits to the generated file itself.Source: Coding guidelines
.agents/skills/nemoclaw-user-reference/references/commands.md (1)
1-1:⚠️ Potential issue | 🟠 Major | ⚡ Quick winAdd the required SPDX license header at the top of this Markdown file.
This file is missing the required SPDX copyright and license header.
🤖 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 @.agents/skills/nemoclaw-user-reference/references/commands.md at line 1, Add the required SPDX header at the very top of this Markdown file above the "# NemoClaw CLI Commands Reference" title by inserting an SPDX header line containing the copyright holder and the license identifier (e.g., include an SPDX-FileCopyrightText entry and an SPDX-License-Identifier line) so the file complies with the project's licensing headers.Source: Coding guidelines
🤖 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.
Inline comments:
In
@.agents/skills/nemoclaw-user-configure-security/references/best-practices.md:
- Around line 237-244: The change modifies an autogenerated skill file
(.agents/skills/nemoclaw-user-configure-security/references/best-practices.md)
which must not be edited directly; instead, update the canonical docs under
docs/ to include the /dev/pts content (or revert your manual edit), then
regenerate the skill YAML/MD using scripts/docs-to-skills.py so the change is
produced automatically; finally replace the manual edit with the regenerated
output and commit that—do not keep direct edits to files matching
.agents/skills/nemoclaw-user-*/*.md.
In `@docs/about/release-notes.mdx`:
- Around line 20-24: The release-notes bullets in docs/about/release-notes.mdx
each contain two sentences on one line; split each bullet so the leading summary
sentence (e.g., the line starting "GPU sandbox creation and local inference
checks now match the runtime paths agents use.", "Onboarding and recovery fail
earlier and stay quieter on common host drift.", "Sandbox state and OpenClaw
operations recover better after direct in-sandbox changes.", "Hermes and
messaging setup expose clearer operator state.", and "Release validation and
maintainability improved with typed E2E scenario fixtures, registry-driven
scenario fan-out, post-reboot recovery guards, latency retry handling, and
aligned CI/test sharding.") is on its own line, and move the explanatory
sentence that follows each summary onto the next line so each bullet uses one
sentence per line for improved diff readability.
In `@docs/inference/use-local-inference.mdx`:
- Around line 214-216: Update the paragraph describing the runtime verification
to state it is conditional: clarify that the onboard runtime probe of
https://inference.local/v1/models is performed when the sandbox runtime has
probe tooling (e.g., curl) available, but onboarding will soft-skip the probe
and continue for minimal images that lack curl (reported as
probe-tool-unavailable) rather than failing; keep the note that only a 2xx
response is treated as success and that, if the probe runs and fails, onboarding
will report the endpoint and recovery steps before the first agent prompt.
---
Outside diff comments:
In @.agents/skills/nemoclaw-user-reference/references/commands.md:
- Line 1: Add the required SPDX header at the very top of this Markdown file
above the "# NemoClaw CLI Commands Reference" title by inserting an SPDX header
line containing the copyright holder and the license identifier (e.g., include
an SPDX-FileCopyrightText entry and an SPDX-License-Identifier line) so the file
complies with the project's licensing headers.
In @.agents/skills/nemoclaw-user-reference/references/network-policies.md:
- Around line 15-24: This autogenerated markdown under the nemoclaw-user skill
was edited directly; instead, update the canonical source docs that feed the
skill generator (e.g., the docs/ entry that documents the filesystem table and
the /dev/pts change), then run the regeneration script scripts/docs-to-skills.py
so the change appears in the generated .agents/skills/nemoclaw-user-*/*.md
files; do not commit manual edits to the generated file itself.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Enterprise
Run ID: c48935b9-24ec-4e32-8770-b9b94043f944
📒 Files selected for processing (10)
.agents/skills/nemoclaw-user-configure-inference/SKILL.md.agents/skills/nemoclaw-user-configure-security/references/best-practices.md.agents/skills/nemoclaw-user-manage-sandboxes/references/backup-restore.md.agents/skills/nemoclaw-user-overview/references/release-notes.md.agents/skills/nemoclaw-user-reference/references/commands.md.agents/skills/nemoclaw-user-reference/references/network-policies.mddocs/about/release-notes.mdxdocs/inference/use-local-inference.mdxdocs/reference/commands-nemohermes.mdxdocs/reference/commands.mdx
Comment on lines
+237
to
244
| The agent has read-write access to `/sandbox`, `/tmp`, `/dev/null`, and `/dev/pts`. | ||
|
|
||
| | Aspect | Detail | | ||
| |---|---| | ||
| | Default | `/sandbox` (agent workspace), `/tmp` (temporary files), `/dev/null`. | | ||
| | Default | `/sandbox` (agent workspace), `/tmp` (temporary files), `/dev/null`, and `/dev/pts` (the devpts pseudo-terminal directory, required so PTY-based tools such as `tmux`, `script`, and interactive shells can allocate a terminal). | | ||
| | What you can change | Add additional writable paths in `filesystem_policy.read_write`. | | ||
| | Risk if relaxed | Each additional writable path expands the agent's ability to persist data and potentially modify system behavior. Adding `/var` lets the agent write to log directories. Adding `/home` gives access to other user directories. | | ||
| | Recommendation | Keep writable paths to `/sandbox` and `/tmp`. If the agent needs a persistent working directory, create a subdirectory under `/sandbox`. | |
Contributor
There was a problem hiding this comment.
Autogenerated skill file should not be edited directly.
Per coding guidelines, files matching .agents/skills/nemoclaw-user-*/*.md are autogenerated and must never be edited directly. The PR objectives mention "Refresh generated NemoClaw user skills from the current docs source" — these changes should originate from the source docs (e.g., docs/) and be regenerated via scripts/docs-to-skills.py, not manually applied to the skill file.
If the source docs have been updated with the /dev/pts content, regenerate the skills. If not, update the source docs first, then regenerate.
Based on coding guidelines: "Autogenerated user skills under .agents/skills/nemoclaw-user-*/*.md must never be edited directly"
🧰 Tools
🪛 LanguageTool
[style] ~243-~243: Try using a synonym here to strengthen your writing.
Context: ...rite to log directories. Adding /home gives access to other user directories. | | R...
(GIVE_PROVIDE)
🤖 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 @.agents/skills/nemoclaw-user-configure-security/references/best-practices.md
around lines 237 - 244, The change modifies an autogenerated skill file
(.agents/skills/nemoclaw-user-configure-security/references/best-practices.md)
which must not be edited directly; instead, update the canonical docs under
docs/ to include the /dev/pts content (or revert your manual edit), then
regenerate the skill YAML/MD using scripts/docs-to-skills.py so the change is
produced automatically; finally replace the manual edit with the regenerated
output and commit that—do not keep direct edits to files matching
.agents/skills/nemoclaw-user-*/*.md.
Source: Coding guidelines
Comment on lines
+214
to
+216
| On Linux Docker-driver GPU sandboxes, NemoClaw keeps local inference on the OpenShell bridge route and verifies `https://inference.local/v1/models` from inside the sandbox runtime after the sandbox reaches ready. | ||
| It treats only a 2xx response as success because that path includes the proxy authentication rewrite the agent uses. | ||
| If the runtime route fails, onboarding reports the endpoint and recovery steps before the first agent prompt. |
Contributor
There was a problem hiding this comment.
Clarify that the runtime probe can be skipped when curl is unavailable.
The text currently reads as unconditional verification, but onboarding can soft-skip this check for minimal images that lack curl (probe-tool-unavailable) and continue. Please make this conditional so operator expectations match behavior.
🤖 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/inference/use-local-inference.mdx` around lines 214 - 216, Update the
paragraph describing the runtime verification to state it is conditional:
clarify that the onboard runtime probe of https://inference.local/v1/models is
performed when the sandbox runtime has probe tooling (e.g., curl) available, but
onboarding will soft-skip the probe and continue for minimal images that lack
curl (reported as probe-tool-unavailable) rather than failing; keep the note
that only a 2xx response is treated as success and that, if the probe runs and
fails, onboarding will report the endpoint and recovery steps before the first
agent prompt.
prekshivyas
approved these changes
Jun 10, 2026
prekshivyas
left a comment
Contributor
There was a problem hiding this comment.
Reviewed (docs accuracy + docs↔skills consistency). Refreshes the v0.0.62 release notes and documents the shipped features across the reference pages: GPU CDI + Jetson/Tegra container recreation, inference.local runtime verification, the nemoclaw <name> agents list passthrough, the Hermes port-8642 dashboard reservation, and /dev/pts becoming writable for PTY tools.
✅ Approve. Verified the load-bearing claims against the code rather than eyeballing:
nemoclaw <name> agents listpassthrough is real (public-display-agents.ts→sandbox:agents:list; the registry test names the add/delete/list trio).- Port
8642is the Hermes API port (gateway/forward port, regression #2078). /dev/ptsis in the read-write baseline (agents/openclaw/policy-permissive.yaml, comment cites the tmux flow).- Ran
scripts/docs-to-skills.pylocally — all five touched.agents/skills/**files match the generated output byte-for-byte, so the docs↔skills mirror is consistent. ($$nemoclawis the expected Fern escape, stripped tonemoclawin the skills output — correct.)
One note, not blocking: static-checks is skipping on this PR, which is the job that runs the "Verify docs-to-skills output" hook — so that consistency check isn't running in CI here. I reproduced it locally and it's clean, and docs-only-checks passed (Fern build valid), so I'm comfortable. Nice, accurate refresh.
jyaunches
approved these changes
Jun 10, 2026
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
4 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.
Summary
Source Summary
docs/reference/commands.mdx: Document CDI-first Docker GPU recreation behavior for Linux Docker-driver sandboxes.docs/inference/use-local-inference.mdx: Document sandbox-runtime verification of theinference.locallocal inference route.docs/reference/commands.mdx: Document Jetson/Tegra device-node group propagation for sandbox CUDA initialization.docs/about/release-notes.mdx: Summarize onboarding and recovery reliability fixes, including the reserved Hermes API port guard.docs/about/release-notes.mdx,docs/reference/commands.mdx: Summarize mutable OpenClaw config recovery and host-sideagents listcoverage.docs/about/release-notes.mdx: Summarize Hermes upstream metadata visibility and WhatsApp QR rendering reliability..agents/skills/: Refresh generated user-skill references from the current docs source.Skipped
openclaw-sandbox-permissive.yaml, which matchesdocs/.docs-skip. Existing source docs remain the source for generated skill synchronization.Verification
python3 scripts/docs-to-skills.py docs/ .agents/skills/ --prefix nemoclaw-user --doc-platform fern-mdxnpm run docs(passes; Fern reports 0 errors and 1 hidden warning)Summary by CodeRabbit
New Features
nemoclaw <name> agents listcommand.Bug Fixes
Documentation
/dev/pts), port 8642 restriction, and command examples.