fix(onboard): surface network-policy hint when OpenClaw plugin install fails during sandbox build by Dongni-Yang · Pull Request #5835 · NVIDIA/NemoClaw

Dongni-Yang · 2026-06-26T04:01:14Z

Summary

When openshell sandbox create fails at openclaw plugins install because npm or ClawHub is unreachable, NemoClaw currently emits only generic resume guidance. This change classifies that narrow failure and prints an actionable network-policy hint while preserving nemoclaw onboard --resume as the recovery path.

Fixes #4127

Changes

Add plugin_install_network_denied to the sandbox-create failure classification.
Require the failed Docker command summary to contain openclaw plugins install.
Require npm-prefixed network evidence to name the same scoped plugin package as the failed command.
Keep non-network failures, later-step failures, different-package errors, package-agnostic errors, and non-plugin npm commands on the generic recovery path.
Add targeted recovery guidance for npm/ClawHub reachability and feature opt-out.
Isolate the nine classifier regressions in src/lib/validation-plugin-install.test.ts.

Source Boundary and Removal Condition

Invalid state: a plugin-install egress failure is surfaced as an undifferentiated sandbox-build failure.
Source boundary: the current OpenShell create interface exposes combined Docker text, including subprocess stderr followed by the failed-command summary, rather than a structured failed-subcommand type.
Source-fix constraint: NemoClaw cannot recover package identity from a structured OpenShell event until that interface exists, and this PR does not patch upstream runtime behavior.
Regression boundary: the classifier requires the exact installer command, same-package npm evidence, and a known network failure; focused positive and false-positive fixtures lock the ordering and quoting assumptions.
Removal condition: replace and remove this text classifier when OpenShell exposes a structured plugin-install failure carrying package identity.

Type of Change

Code change (bug fix)
Code change with doc updates
Doc only

Quality Gates

Focused tests cover the changed behavior and false-positive boundaries.
No user documentation change is needed; the emitted recovery hint is self-contained.
The code is diagnostic only and does not widen policy or change sandbox state.
Live runtime validation is not required for this parser-only correction: deterministically forcing external registry denial would change the environment, while the pure classifier and printer are covered with Docker/OpenShell-shaped fixtures.

Verification

PR body includes the DCO declaration and every commit appears Verified in GitHub.
Commit and push hooks passed.
npx vitest run --project cli src/lib/validation.test.ts src/lib/validation-plugin-install.test.ts src/lib/build-context.test.ts passed: 87 tests.
npm run typecheck:cli passed.
npm run test:imports:check passed.
Full root CLI/integration hook suite passed.
No secrets, API keys, or credentials committed.

Signed-off-by: Dongni Yang dongniy@nvidia.com

…the openclaw plugins install step When `openshell sandbox create` fails because the Dockerfile's `openclaw plugins install` RUN step exits non-zero, the build output contains the distinctive strings "openclaw plugins install" or "npm:@openclaw/", but `classifySandboxCreateFailure` had no branch for this pattern and fell through to `kind="unknown"`. The resulting hint was the generic "nemoclaw onboard --resume" line with no indication of the likely cause. Add a `"plugin_install_network_denied"` kind to `SandboxCreateFailure` and a classifier that matches those strings in the openshell create output. `printSandboxCreateRecoveryHints` now emits a targeted message noting that network policy may be blocking outbound access to ClawHub or the npm registry, and suggests disabling the feature (e.g. `NEMOCLAW_WEB_SEARCH_ENABLED=0`) as an alternative if a preset isn't available. Fixes #4127 Signed-off-by: Dongni Yang <dongniy@nvidia.com>

The previous regex matched `openclaw plugins install|npm:@openclaw/` anywhere in the captured output, which would mis-classify a build where that step succeeded and a later step failed — the Docker step header (`Step N: RUN openclaw plugins install ...`) would match even though the failure came from an unrelated subsequent RUN step. Tighten to the Docker error block format: The command '...<plugin text>...' returned a non-zero code `[^']*` is used (not `[^\n]*`) because the embedded shell command often spans multiple lines (chained commands joined with semicolons), and JS character classes match newlines. Add a regression test that verifies the step-header false-positive is rejected. Refs #4127 Signed-off-by: Dongni Yang <dongniy@nvidia.com>

…n-install failure as network-denied The previous classifier fired on any failed `openclaw plugins install` Docker command, including package-not-found (HTTP 404 from the registry), version conflicts, and auth errors — all of which would receive a misleading "network policy blocking egress" hint. Narrow the match to require both: 1. The Docker error block anchored to the failed plugin-install step 2. A network/egress reachability error (ENOTFOUND, EAI_AGAIN, ECONNREFUSED, ETIMEDOUT, fetch failed, etc.) in the captured output This ensures the hint "your sandbox network policy may be blocking outbound plugin-install access" is only shown when the underlying failure is actually a network reachability problem. Test changes: - Updated positive tests to include ENOTFOUND / ECONNREFUSED output matching real npm network error messages (registry.npmjs.org and ClawHub paths respectively) - Added negative test: same Docker failed-command block but E404 package-not-found → classifies as "unknown" - Added direct build-context.test.ts coverage for the new hint branch, asserting the key user-visible strings (plugin-install step, ClawHub, npm registry, network policy, NEMOCLAW_WEB_SEARCH_ENABLED=0, onboard --resume) Refs #4127 Signed-off-by: Dongni Yang <dongniy@nvidia.com>

coderabbitai · 2026-06-26T04:01:36Z

📝 Walkthrough

Walkthrough

Adds a new sandbox-create failure kind for plugin-install network denial, classifies matching OpenClaw install errors, and prints recovery hints that mention the plugin-install step, network policy, and onboard --resume.

Changes

OpenClaw plugin install network denial

Layer / File(s)	Summary
Failure kind and classifier `src/lib/validation.ts`, `src/lib/validation.test.ts`	Adds `plugin_install_network_denied` to sandbox-create failures and classifies plugin-install failures when network/DNS-style error text is present; tests cover matching and non-matching cases.
Recovery hints and resume path `src/lib/build-context.ts`, `src/lib/build-context.test.ts`	Prints targeted recovery guidance for the new failure kind, including the plugin-install step context and `onboard --resume`, with a test for the emitted text.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Possibly related issues

[NemoClaw][All Platforms] Default sandbox network policy blocks clawhub, preventing openclaw plugins install from succeeding #4104: The new classifier and recovery hints address the same OpenClaw plugin-install reachability failures involving blocked ClawHub/npm access.

Suggested labels

integration: openclaw, area: onboarding, area: sandbox, bug-fix

Suggested reviewers

cv
ericksoa

Poem

A rabbit hopped through logs so bright,
and found a network-denied night.
“ClawHub’s far,” said I with flair,
“but onboard --resume gets us there!”
🐇✨

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly summarizes the main change: a network-policy hint for OpenClaw plugin-install sandbox build failures.
Linked Issues check	✅ Passed	The PR adds targeted diagnostics for plugin-install egress denials and covers both npm registry and ClawHub network failures as requested.
Out of Scope Changes check	✅ Passed	No unrelated functionality is introduced beyond the plugin-install failure diagnostics, classifier, and tests.
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.

✨ Finishing Touches

📝 Generate docstrings

Create stacked PR
Commit on current branch

🧪 Generate unit tests (beta)

Create PR with unit tests
Commit unit tests in branch fix/4127-plugin-install-network-policy-hint

_{Comment @coderabbitai help to get the list of available commands.}

github-code-quality · 2026-06-26T04:01:55Z

Code Coverage Overview

Languages: TypeScript

TypeScript / code-coverage/plugin

The overall coverage in the fix/4127-plugin-inst... branch is 96%. Coverage data for the main branch is not yet available.

Show a code coverage summary of the most covered files.

File	main	fix/4127-plugin-inst... `8d4ee7e`	+/-
`nemoclaw/src/se...cret-scanner.ts`	—	100%	—
`nemoclaw/src/commands/slash.ts`	—	100%	—
`nemoclaw/src/li...bprocess-env.ts`	—	100%	—
`nemoclaw/src/bl...eprint/state.ts`	—	98%	—
`nemoclaw/src/onboard/config.ts`	—	98%	—
`nemoclaw/src/bl...int/snapshot.ts`	—	97%	—
`nemoclaw/src/bl...print/runner.ts`	—	95%	—
`nemoclaw/src/co...ration-state.ts`	—	94%	—
`nemoclaw/src/bl...ate-networks.ts`	—	94%	—
`nemoclaw/src/index.ts`	—	94%	—

TypeScript / code-coverage/cli

The overall coverage in the fix/4127-plugin-inst... branch is 67%. Coverage data for the main branch is not yet available.

Show a code coverage summary of the most covered files.

File	main	fix/4127-plugin-inst... `8d4ee7e`	+/-
`src/lib/actions...all/run-plan.ts`	—	80%	—
`src/lib/state/o...oard-session.ts`	—	79%	—
`src/lib/actions...dbox/rebuild.ts`	—	74%	—
`src/lib/state/sandbox.ts`	—	72%	—
`src/lib/onboard/preflight.ts`	—	69%	—
`src/lib/actions...licy-channel.ts`	—	60%	—
`src/lib/onboard...er-gpu-patch.ts`	—	59%	—
`src/lib/policy/index.ts`	—	52%	—
`src/lib/shields/index.ts`	—	51%	—
`src/lib/onboard.ts`	—	20%	—

_{Updated June 30, 2026 05:17 UTC
Code Coverage is in Public Preview. Learn more and provide us with your feedback.}

github-actions · 2026-06-26T04:03:49Z

PR Review Advisor — No blocking findings

Merge posture: No blocking advisor findings
Primary next action: No advisor follow-up required beyond maintainer review.
Open items: 0 required · 0 warnings · 0 suggestions · 0 test follow-ups
Top item: No actionable code findings

Workflow run details

This is an automated, non-binding review; it still expects maintainers and agents to respond to each required or warning item. Treat suggestions as current-PR improvements when they touch changed code; defer only with maintainer rationale or a linked follow-up. A human maintainer must make the final merge decision.

github-actions · 2026-06-26T04:03:52Z

E2E Advisor Recommendation

Required E2E: onboard-negative-paths, brave-search
Optional E2E: network-policy

Dispatch hint: onboard-negative-paths,brave-search

Workflow run

Full advisor summary

E2E Recommendation Advisor

Base: origin/main
Head: HEAD
Confidence: high

Required E2E

onboard-negative-paths (low): Closest existing negative-path E2E for onboarding CLI failure/output behavior; this PR changes failure classification and recovery text users see from onboarding failures.
brave-search (medium): Exercises the real OpenClaw plugin-install path used by the web-search plugin flow, which is the concrete plugin-install scenario referenced by the new classifier and recovery hint.

Optional E2E

network-policy (high): Useful adjacent confidence because the new hint points users at sandbox network-policy egress for npm/ClawHub, though this PR does not modify policy assets or enforcement.

New E2E recommendations

plugin-install sandbox-create failure recovery (high): Existing E2E jobs do not appear to deterministically inject an npm/ClawHub network failure at the openclaw plugins install Docker build step and assert the new plugin_install_network_denied recovery guidance.
- Suggested test: Add a focused live or regression E2E that forces the sandbox Docker build plugin-install step to fail with npm ENOTFOUND/ECONNREFUSED for the selected @openclaw package, then asserts the CLI output mentions OpenClaw plugin-install, ClawHub/npm registry, network policy, NEMOCLAW_WEB_SEARCH_ENABLED=0, and onboard --resume.

Dispatch hint

Workflow: .github/workflows/e2e.yaml
jobs input: onboard-negative-paths,brave-search

github-actions · 2026-06-26T04:03:52Z

Vitest E2E Scenario Recommendation

Required Vitest E2E scenarios: None
Optional Vitest E2E scenarios: None

Workflow run

Full Vitest E2E advisor summary

Vitest E2E Scenario Advisor

Base: origin/main
Head: HEAD
Confidence: high

Required Vitest E2E scenarios

None. Changes are limited to sandbox-create failure classification and user-facing recovery hints for a plugin-install network-denial error, with accompanying unit tests. The live Vitest scenario workflow does not have a deterministic typed scenario that forces the OpenClaw plugin-install network-policy failure path, so a scenario dispatch would not prove this changed surface.

Optional Vitest E2E scenarios

None.

Relevant changed files

src/lib/build-context.ts
src/lib/validation.ts

coderabbitai

Actionable comments posted: 1

🤖 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 `@src/lib/validation.ts`:
- Around line 144-149: The classifier in validation.ts is too broad because it
matches any Docker RUN block containing openclaw plugins install, even when a
later command in the same block fails. Tighten the logic in the network-failure
branch so it only classifies plugin_install_network_denied when the failing
subcommand itself shows plugin-fetch/install evidence, using the existing
validation.ts matching and the relevant openclaw plugins install / npm:`@openclaw`
patterns as anchors. Update validation.test.ts with a regression case where
plugin install succeeds but a later command in the same shell block fails, and
verify it no longer returns the plugin install recovery hint.

🪄 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: b142153c-913c-41f6-bb87-87ba3729f360

📥 Commits

Reviewing files that changed from the base of the PR and between 8e047a6 and 563708c.

📒 Files selected for processing (4)

src/lib/build-context.test.ts
src/lib/build-context.ts
src/lib/validation.test.ts
src/lib/validation.ts

…false positives from later RUN block commands The previous network-error predicate matched any ENOTFOUND/ECONNREFUSED/etc. in the captured output, which could fire the plugin-install hint when the plugin install itself succeeded but a later command in the same RUN block (e.g. `openclaw doctor --fix`) failed with a network error. npm's error output consistently prefixes every line with "npm error", whereas commands run after a successful install produce different error formats. Requiring "npm error" + network pattern anchors the evidence to the npm installer specifically, ruling out later-command false positives. Add a regression test: the same RUN block runs plugin install (succeeds) then openclaw doctor (fails with ENOTFOUND but no "npm error" prefix) → correctly returns "unknown". Refs #4127 Signed-off-by: Dongni Yang <dongniy@nvidia.com>

…get divergence Refs #4127 Signed-off-by: Dongni Yang <dongniy@nvidia.com>

Bound the network-evidence regex to the text up to and including the failed-plugin-install Docker error block rather than scanning the entire output. This prevents an npm script that runs after a successful plugin install in the same RUN block from producing a false plugin_install_network_denied classification when that later script emits an npm-prefixed network error. Adds a regression test for this case (npm script post-install in the same RUN block emits npm error ENOTFOUND after the Docker boundary). Refs #4127 Signed-off-by: Dongni Yang <dongniy@nvidia.com>

Dongni-Yang · 2026-06-26T06:59:57Z

Addressing PRA-2 / CodeRabbit (windowed npm-error search):

Both PRA-2 and the CodeRabbit comment correctly identified that scanning the entire output for npm error network evidence could misfire if an unrelated npm script in the same RUN block emits a network error after the plugin-install step has already succeeded.

Fixed in the latest commit (9d8df9496): the network-evidence regex now searches only the text up to and including the Docker The command '...' returned a non-zero code boundary, not the full output. This means npm errors that appear after the Docker error line (i.e. from a later command whose errors are streamed after Docker reports the block failure) are excluded from the classification window.

A regression test was added for this case:

"The command '...openclaw plugins install...' returned a non-zero code: 1"
"npm error code ENOTFOUND"       ← appears after the Docker boundary → excluded
"npm error network request ..."  ← same
→ kind: "unknown"  ✓

Addressing PRA-1 (source-of-truth review):

The classification is intentionally localized to NemoClaw's classifier layer. The source-of-truth facts:

Invalid state: Docker build fails at the plugin-install RUN block with both (a) a The command '...' returned a non-zero code boundary containing openclaw plugins install or npm:@openclaw/, and (b) an npm error network error in the prefix up to that boundary.
Source boundary: classifySandboxCreateFailure() in src/lib/validation.ts — a pure function with no I/O, so the behavior is fully unit-testable without Docker.
Source-fix constraint: The upstream fix would be OpenShell exposing the failing RUN-block step name in structured error output so the classifier doesn't need regex parsing. Until that lands, the localized pattern match is the best available signal.
Regression tests: 7 tests total covering the positive path (ENOTFOUND, ECONNREFUSED against npm/ClawHub registries), 5 negative paths (unrelated failures, non-network E404, step-header false positive, doctor-fix format, post-boundary npm error).
Removal condition: when OpenShell exposes a structured failedStep field in sandbox create failure output, this classifier branch can be replaced with a direct field check.

github-actions · 2026-06-26T20:05:25Z

Selective E2E Results — ✅ All requested jobs passed

Run: 28261980087
Target ref: fix/4127-plugin-install-network-policy-hint
Requested jobs: onboard-negative-paths-e2e,cloud-onboard-e2e
Summary: 2 passed, 0 failed, 0 cancelled, 0 skipped

Job	Result
cloud-onboard-e2e	✅ success
onboard-negative-paths-e2e	✅ success

prekshivyas

CodeRabbit's concern about same-RUN-block false positives is pre-addressed: the window slice text.slice(0, match.index + match[0].length) excludes post-boundary npm output, and the npm error.* prefix anchor drops non-npm error formats. Tests at lines 344 and 359 of validation.test.ts cover both sub-cases (non-npm network error and npm network error from a later command in the same block). Implementation is correct as-is.

github-actions · 2026-06-30T04:24:49Z

E2E Target Recommendation

Required E2E targets: ubuntu-repo-cloud-openclaw
Optional E2E targets: None

Dispatch required E2E targets:

gh workflow run e2e.yaml --ref <pr-head-ref> --field targets=ubuntu-repo-cloud-openclaw

Workflow run

Full E2E target advisor summary

E2E Target Advisor

Base: origin/main
Head: HEAD
Confidence: medium

Required E2E targets

ubuntu-repo-cloud-openclaw: The PR changes sandbox-create failure classification and recovery hints used by OpenClaw onboarding. The smallest live-supported typed target that exercises the OpenClaw Docker onboarding and sandbox creation surface is ubuntu-repo-cloud-openclaw.
- Dispatch: gh workflow run e2e.yaml --ref <pr-head-ref> --field targets=ubuntu-repo-cloud-openclaw

Optional E2E targets

None.

Relevant changed files

src/lib/build-context.ts
src/lib/validation.ts

github-actions · 2026-06-30T04:26:10Z

PR Review Advisor (Nemotron Ultra) — No blocking findings

Merge posture: No blocking advisor findings
Primary next action: Add or justify PRA-T1 and any related test follow-ups.
Open items: 0 required · 0 warnings · 2 suggestions · 1 test follow-up
Since last review: 0 prior items resolved · 0 still apply · 0 new items found

Action checklist

PRA-T1 Add or justify test follow-up: Test uses hardcoded plugin version that will age
PRA-1 In-scope improvement: Test uses hardcoded plugin version that will age in src/lib/validation-plugin-install.test.ts:15
PRA-2 In-scope improvement: Hint references single feature flag as example; consider generalizing if more plugin-dependent features exist in src/lib/build-context.ts:231

Findings index

ID	Severity	Category	Location	Required action
`PRA-1`	Improvement	tests	`src/lib/validation-plugin-install.test.ts:15`	Replace with a generic version like `@latest` or `@1.0.0` to signal version is irrelevant, or add a comment that version is ignored.
`PRA-2`	Improvement	scope	`src/lib/build-context.ts:231`	Add a TODO comment noting that if more plugin-dependent features are added, the hint should generalize (e.g., 'disable the feature requiring this plugin').

Review findings by urgency: 0 required fixes, 0 items to resolve/justify, 2 in-scope improvements

⚠️ Resolve or justify before merge

Investigate these in the current review; either fix them, explain why they are not applicable, or document the accepted risk.

None.

💡 In-scope improvements

These are lower-risk, not throwaway. Prefer fixing them in this PR when they are local to changed code; defer only with rationale or a linked follow-up.

`PRA-1` Improvement — Test uses hardcoded plugin version that will age

Location: src/lib/validation-plugin-install.test.ts:15
Category: tests
Problem: The positive test case embeds a specific version `@2026.5.27` in the Docker command text. While the classifier only matches the package name (not version), using a fixed date-like version could confuse future readers into thinking the version matters.
Impact: Low — cosmetic; no functional risk since classifier extracts package name via regex ignoring version.
Suggested action: Replace with a generic version like `@latest` or `@1.0.0` to signal version is irrelevant, or add a comment that version is ignored.
Expected follow-up: Prefer a current-PR fix when local to changed code; defer only with rationale or linked follow-up.
Verification: Read test line 15; confirm classifier regex at validation.ts:147 only captures @openclaw/... package name.
Missing regression test: No new test needed — existing test already proves version independence via package-name extraction logic.
Done when: The local improvement is applied, or the PR notes why it should be deferred.
Evidence: validation-plugin-install.test.ts:15 contains 'npm:@openclaw/brave-plugin@2026.5.27'; validation.ts:147 regex /(?:npm:)?(@openclaw\/[a-z0-9._-]+)/gi captures only package name.

`PRA-2` Improvement — Hint references single feature flag as example; consider generalizing if more plugin-dependent features exist

Location: src/lib/build-context.ts:231
Category: scope
Problem: Recovery hint mentions `NEMOCLAW_WEB_SEARCH_ENABLED=0` as the example feature to disable. If additional features trigger plugin installs, the hint may need updating.
Impact: Low — hint is guidance only; users can still inspect policy. No functional regression if example becomes stale.
Suggested action: Add a TODO comment noting that if more plugin-dependent features are added, the hint should generalize (e.g., 'disable the feature requiring this plugin').
Expected follow-up: Prefer a current-PR fix when local to changed code; defer only with rationale or linked follow-up.
Verification: Read build-context.ts:221-233; check if other env flags gate plugin installs in codebase.
Missing regression test: Not testable — depends on future feature additions.
Done when: The local improvement is applied, or the PR notes why it should be deferred.
Evidence: build-context.ts:231 mentions NEMOCLAW_WEB_SEARCH_ENABLED=0; grep for other NEMOCLAW_*_ENABLED flags that might gate plugins.

Test follow-ups to resolve or justify

If these cover changed behavior, prefer adding them in this PR; otherwise state why existing coverage is enough or link the follow-up.

PRA-T1 Test uses hardcoded plugin version that will age — Replace with a generic version like `@latest` or `@1.0.0` to signal version is irrelevant, or add a comment that version is ignored.

Since last review details

Current findings, using the urgency labels above:

`PRA-1` Improvement — Test uses hardcoded plugin version that will age

Location: src/lib/validation-plugin-install.test.ts:15
Category: tests
Problem: The positive test case embeds a specific version `@2026.5.27` in the Docker command text. While the classifier only matches the package name (not version), using a fixed date-like version could confuse future readers into thinking the version matters.
Impact: Low — cosmetic; no functional risk since classifier extracts package name via regex ignoring version.
Suggested action: Replace with a generic version like `@latest` or `@1.0.0` to signal version is irrelevant, or add a comment that version is ignored.
Expected follow-up: Prefer a current-PR fix when local to changed code; defer only with rationale or linked follow-up.
Verification: Read test line 15; confirm classifier regex at validation.ts:147 only captures @openclaw/... package name.
Missing regression test: No new test needed — existing test already proves version independence via package-name extraction logic.
Done when: The local improvement is applied, or the PR notes why it should be deferred.
Evidence: validation-plugin-install.test.ts:15 contains 'npm:@openclaw/brave-plugin@2026.5.27'; validation.ts:147 regex /(?:npm:)?(@openclaw\/[a-z0-9._-]+)/gi captures only package name.

`PRA-2` Improvement — Hint references single feature flag as example; consider generalizing if more plugin-dependent features exist

Location: src/lib/build-context.ts:231
Category: scope
Problem: Recovery hint mentions `NEMOCLAW_WEB_SEARCH_ENABLED=0` as the example feature to disable. If additional features trigger plugin installs, the hint may need updating.
Impact: Low — hint is guidance only; users can still inspect policy. No functional regression if example becomes stale.
Suggested action: Add a TODO comment noting that if more plugin-dependent features are added, the hint should generalize (e.g., 'disable the feature requiring this plugin').
Expected follow-up: Prefer a current-PR fix when local to changed code; defer only with rationale or linked follow-up.
Verification: Read build-context.ts:221-233; check if other env flags gate plugin installs in codebase.
Missing regression test: Not testable — depends on future feature additions.
Done when: The local improvement is applied, or the PR notes why it should be deferred.
Evidence: build-context.ts:231 mentions NEMOCLAW_WEB_SEARCH_ENABLED=0; grep for other NEMOCLAW_*_ENABLED flags that might gate plugins.

Workflow run details

This is an automated, non-binding review; it still expects maintainers and agents to respond to each required or warning item. Treat suggestions as current-PR improvements when they touch changed code; defer only with maintainer rationale or a linked follow-up. A human maintainer must make the final merge decision.

Signed-off-by: Carlos Villela <cvillela@nvidia.com>

…ll-network-policy-hint

Signed-off-by: Carlos Villela <cvillela@nvidia.com>

cv · 2026-06-30T06:12:37Z

Maintainer disposition for exact head 8d4ee7e3715fdd7851aedc80dd2551fbb1c049e7:

Both PR Review Advisor jobs/checks are green, and the workflow is explicitly advisory. Nemotron PRA-1 is internally contradictory: its own required action says “No action needed” and cites the committed source-boundary/removal-condition comment plus nine focused fixtures. Treat PRA-1 as resolved.
Accept the bounded text-parser risk in PRA-2 and PRA-T1–PRA-T3: package correlation, Docker-boundary slicing, network-code matching, and negative fixtures cover the changed behavior. A live success-path onboard target cannot exercise this failure classifier.
PRA-3 and PRA-4 concern unchanged code and are out of scope.
PRA-5/PRA-T4 does not match the production installer shape, which invokes openclaw plugins install once per package.
PRA-6 is mitigated by same-package correlation and the package-agnostic negative test. PRA-7 proposes a broad recovery-plan refactor; this PR directly tests the existing hint-rendering pattern.

No code change or additional E2E is required for these advisor items. Rerunning the PR Review Advisor once on this exact SHA to refresh its sticky outputs.

cv · 2026-06-30T06:19:09Z

Refreshed Nemotron suggestions disposition: keep @2026.5.27 in the fixture because it intentionally mirrors the pinned OpenClaw runtime output; the classifier ignores the version. No TODO is needed for the feature flag because the emitted hint is already generic (“the feature that requires this plugin”) and labels NEMOCLAW_WEB_SEARCH_ENABLED=0 only as an example. These are non-blocking and require no code change.

## Summary Refreshes the public documentation for NemoClaw v0.0.71 after scanning commits since v0.0.70. Adds release notes and fills the remaining doc gaps for Windows bootstrap diagnostics, OpenClaw agent auto-relock warnings, auto-pair cadence tuning, and plugin-install recovery hints. ## Changes - `docs/about/release-notes.mdx`: adds the v0.0.71 release-note section, grouped by gateway recovery, OpenShell auth, policy provenance, day-two maintenance, messaging/inference, and Windows setup. - `docs/get-started/windows-preparation.mdx`: documents sanitized WSL install output and reboot gating in the Windows bootstrap. - `docs/reference/commands.mdx`: documents the host `agent` wrapper's shields auto-relock warning and OpenClaw auto-pair watcher tuning variables. - `docs/reference/troubleshooting.mdx`: adds plugin-install network failure recovery guidance and updates Windows WSL troubleshooting for sanitized install logs and reboot-required handling. Source summary: - #6065 -> `docs/about/release-notes.mdx`: Notes explicit model override preservation and gateway-log guard-chain recovery diagnostics. - #5874 -> `docs/about/release-notes.mdx`: Summarizes host-mediated `recover` and `gateway restart`, linking to lifecycle, command, troubleshooting, and trusted-boundary docs already added by the source PR. - #5596 -> `docs/about/release-notes.mdx`: Summarizes OpenShell 0.0.71 gateway auth, loopback binding, and compatibility-container docs already added by the source PR. - #5797 and #5798 -> `docs/about/release-notes.mdx`: Summarizes `policy-list` provenance, Restricted tier suppression, and Balanced tier weather behavior already reflected in policy docs. - #5784 -> `docs/about/release-notes.mdx`: Summarizes `--destroy-user-data` and the safe `--yes` uninstall behavior already documented in lifecycle and command docs. - #6034 -> `docs/about/release-notes.mdx`: Summarizes custom Dockerfile warm-build cache behavior already documented in the command reference. - #5951 -> `docs/reference/commands.mdx`: Documents the stderr-only host `agent` wrapper warning after recent shields auto-relock. - #5387 -> `docs/reference/commands.mdx`: Documents OpenClaw auto-pair watcher cadence and fast-reentry tuning variables. - #5835 -> `docs/reference/troubleshooting.mdx`: Adds recovery guidance for OpenClaw plugin-install network failures. - #5995 and #5956 -> `docs/about/release-notes.mdx`: Summarizes Microsoft Teams final-message delivery and runtime mention hints already covered by messaging docs. - #5716 -> `docs/about/release-notes.mdx`: Summarizes non-interactive Ollama loopback safety already covered by local inference docs. - #5505, #5527, and #5528 -> `docs/about/release-notes.mdx`: Summarizes compatible local endpoint, model task-fit, and model capability audit docs. - #6009 -> `docs/get-started/windows-preparation.mdx`, `docs/reference/troubleshooting.mdx`: Documents sanitized Windows bootstrap WSL output and reboot-required gating. - #6055 -> no additional source doc page change needed beyond the already-merged quickstart update; release notes did not duplicate routine quickstart cleanup. No matching v0.0.71 GitHub announcement discussion was found in the latest 20 discussions, so this refresh is based on the commit scan and existing source PR docs. ## Type of Change - [ ] Code change (feature, bug fix, or refactor) - [ ] Code change with doc updates - [ ] Doc only (prose changes, no code sample modifications) - [x] Doc only (includes code sample changes) ## Quality Gates  - [ ] Tests added or updated for changed behavior - [ ] Existing tests cover changed behavior — justification: - [x] Tests not applicable — justification: docs-only refresh with no runtime behavior changes. - [x] Docs updated for user-facing behavior changes - [ ] Docs not applicable — justification: - [ ] Sensitive paths changed (security, policy, credentials, preflight, onboarding, inference, runner, sandbox, or messaging) - [ ] Sensitive-path review completed or maintainer-approved waiver recorded — reviewer/approval link/justification: - [ ] Non-success, skipped, or missing CI check accepted by maintainer — check name, approval link, and follow-up issue: ## Verification  - [x] PR description includes the DCO sign-off declaration and every commit appears as `Verified` in GitHub - [x] Git hooks passed during commit and push, or `npx prek run --from-ref main --to-ref HEAD` passes - [ ] Targeted tests pass for changed behavior - [ ] Full `npm test` passes (broad runtime changes only) - [x] Quality Gates section completed with required justifications or waivers - [x] No secrets, API keys, or credentials committed - [ ] `npm run docs` builds without warnings (doc changes only) — ran `npm run docs`; Fern reported 0 errors and 2 existing warnings. - [x] 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) ---  Signed-off-by: Julie Yaunches <jyaunches@nvidia.com>  ## Summary by CodeRabbit * **Documentation** * Added a new release-notes entry covering gateway recovery, authentication, network policy/provenance output, uninstall safety, Windows bootstrap diagnostics, messaging defaults, and inference setup guidance. * Clarified Windows preparation steps around reboot behavior and redacting troubleshooting transcripts. * Expanded command reference details for OpenClaw wrapper behavior and new auto-pair tuning options. * Improved troubleshooting guidance for plugin installation issues, WSL repair/reboot cases, and install timing problems.  Signed-off-by: Julie Yaunches <jyaunches@nvidia.com> Co-authored-by: Prekshi Vyas <34834085+prekshivyas@users.noreply.github.com>

…l fails during sandbox build (NVIDIA#5835)  ## Summary When `openshell sandbox create` fails at `openclaw plugins install` because npm or ClawHub is unreachable, NemoClaw currently emits only generic resume guidance. This change classifies that narrow failure and prints an actionable network-policy hint while preserving `nemoclaw onboard --resume` as the recovery path. Fixes NVIDIA#4127 ## Changes - Add `plugin_install_network_denied` to the sandbox-create failure classification. - Require the failed Docker command summary to contain `openclaw plugins install`. - Require npm-prefixed network evidence to name the same scoped plugin package as the failed command. - Keep non-network failures, later-step failures, different-package errors, package-agnostic errors, and non-plugin npm commands on the generic recovery path. - Add targeted recovery guidance for npm/ClawHub reachability and feature opt-out. - Isolate the nine classifier regressions in `src/lib/validation-plugin-install.test.ts`. ## Source Boundary and Removal Condition - Invalid state: a plugin-install egress failure is surfaced as an undifferentiated sandbox-build failure. - Source boundary: the current OpenShell create interface exposes combined Docker text, including subprocess stderr followed by the failed-command summary, rather than a structured failed-subcommand type. - Source-fix constraint: NemoClaw cannot recover package identity from a structured OpenShell event until that interface exists, and this PR does not patch upstream runtime behavior. - Regression boundary: the classifier requires the exact installer command, same-package npm evidence, and a known network failure; focused positive and false-positive fixtures lock the ordering and quoting assumptions. - Removal condition: replace and remove this text classifier when OpenShell exposes a structured plugin-install failure carrying package identity. ## Type of Change - [x] Code change (bug fix) - [ ] Code change with doc updates - [ ] Doc only ## Quality Gates - [x] Focused tests cover the changed behavior and false-positive boundaries. - [x] No user documentation change is needed; the emitted recovery hint is self-contained. - [x] The code is diagnostic only and does not widen policy or change sandbox state. - [x] Live runtime validation is not required for this parser-only correction: deterministically forcing external registry denial would change the environment, while the pure classifier and printer are covered with Docker/OpenShell-shaped fixtures. ## Verification - [x] PR body includes the DCO declaration and every commit appears Verified in GitHub. - [x] Commit and push hooks passed. - [x] `npx vitest run --project cli src/lib/validation.test.ts src/lib/validation-plugin-install.test.ts src/lib/build-context.test.ts` passed: 87 tests. - [x] `npm run typecheck:cli` passed. - [x] `npm run test:imports:check` passed. - [x] Full root CLI/integration hook suite passed. - [x] No secrets, API keys, or credentials committed. --- Signed-off-by: Dongni Yang <dongniy@nvidia.com> --------- Signed-off-by: Dongni Yang <dongniy@nvidia.com> Signed-off-by: Carlos Villela <cvillela@nvidia.com> Co-authored-by: Carlos Villela <cvillela@nvidia.com>

## Summary Refreshes the public documentation for NemoClaw v0.0.71 after scanning commits since v0.0.70. Adds release notes and fills the remaining doc gaps for Windows bootstrap diagnostics, OpenClaw agent auto-relock warnings, auto-pair cadence tuning, and plugin-install recovery hints. ## Changes - `docs/about/release-notes.mdx`: adds the v0.0.71 release-note section, grouped by gateway recovery, OpenShell auth, policy provenance, day-two maintenance, messaging/inference, and Windows setup. - `docs/get-started/windows-preparation.mdx`: documents sanitized WSL install output and reboot gating in the Windows bootstrap. - `docs/reference/commands.mdx`: documents the host `agent` wrapper's shields auto-relock warning and OpenClaw auto-pair watcher tuning variables. - `docs/reference/troubleshooting.mdx`: adds plugin-install network failure recovery guidance and updates Windows WSL troubleshooting for sanitized install logs and reboot-required handling. Source summary: - NVIDIA#6065 -> `docs/about/release-notes.mdx`: Notes explicit model override preservation and gateway-log guard-chain recovery diagnostics. - NVIDIA#5874 -> `docs/about/release-notes.mdx`: Summarizes host-mediated `recover` and `gateway restart`, linking to lifecycle, command, troubleshooting, and trusted-boundary docs already added by the source PR. - NVIDIA#5596 -> `docs/about/release-notes.mdx`: Summarizes OpenShell 0.0.71 gateway auth, loopback binding, and compatibility-container docs already added by the source PR. - NVIDIA#5797 and NVIDIA#5798 -> `docs/about/release-notes.mdx`: Summarizes `policy-list` provenance, Restricted tier suppression, and Balanced tier weather behavior already reflected in policy docs. - NVIDIA#5784 -> `docs/about/release-notes.mdx`: Summarizes `--destroy-user-data` and the safe `--yes` uninstall behavior already documented in lifecycle and command docs. - NVIDIA#6034 -> `docs/about/release-notes.mdx`: Summarizes custom Dockerfile warm-build cache behavior already documented in the command reference. - NVIDIA#5951 -> `docs/reference/commands.mdx`: Documents the stderr-only host `agent` wrapper warning after recent shields auto-relock. - NVIDIA#5387 -> `docs/reference/commands.mdx`: Documents OpenClaw auto-pair watcher cadence and fast-reentry tuning variables. - NVIDIA#5835 -> `docs/reference/troubleshooting.mdx`: Adds recovery guidance for OpenClaw plugin-install network failures. - NVIDIA#5995 and NVIDIA#5956 -> `docs/about/release-notes.mdx`: Summarizes Microsoft Teams final-message delivery and runtime mention hints already covered by messaging docs. - NVIDIA#5716 -> `docs/about/release-notes.mdx`: Summarizes non-interactive Ollama loopback safety already covered by local inference docs. - NVIDIA#5505, NVIDIA#5527, and NVIDIA#5528 -> `docs/about/release-notes.mdx`: Summarizes compatible local endpoint, model task-fit, and model capability audit docs. - NVIDIA#6009 -> `docs/get-started/windows-preparation.mdx`, `docs/reference/troubleshooting.mdx`: Documents sanitized Windows bootstrap WSL output and reboot-required gating. - NVIDIA#6055 -> no additional source doc page change needed beyond the already-merged quickstart update; release notes did not duplicate routine quickstart cleanup. No matching v0.0.71 GitHub announcement discussion was found in the latest 20 discussions, so this refresh is based on the commit scan and existing source PR docs. ## Type of Change - [ ] Code change (feature, bug fix, or refactor) - [ ] Code change with doc updates - [ ] Doc only (prose changes, no code sample modifications) - [x] Doc only (includes code sample changes) ## Quality Gates  - [ ] Tests added or updated for changed behavior - [ ] Existing tests cover changed behavior — justification: - [x] Tests not applicable — justification: docs-only refresh with no runtime behavior changes. - [x] Docs updated for user-facing behavior changes - [ ] Docs not applicable — justification: - [ ] Sensitive paths changed (security, policy, credentials, preflight, onboarding, inference, runner, sandbox, or messaging) - [ ] Sensitive-path review completed or maintainer-approved waiver recorded — reviewer/approval link/justification: - [ ] Non-success, skipped, or missing CI check accepted by maintainer — check name, approval link, and follow-up issue: ## Verification  - [x] PR description includes the DCO sign-off declaration and every commit appears as `Verified` in GitHub - [x] Git hooks passed during commit and push, or `npx prek run --from-ref main --to-ref HEAD` passes - [ ] Targeted tests pass for changed behavior - [ ] Full `npm test` passes (broad runtime changes only) - [x] Quality Gates section completed with required justifications or waivers - [x] No secrets, API keys, or credentials committed - [ ] `npm run docs` builds without warnings (doc changes only) — ran `npm run docs`; Fern reported 0 errors and 2 existing warnings. - [x] 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) ---  Signed-off-by: Julie Yaunches <jyaunches@nvidia.com>  ## Summary by CodeRabbit * **Documentation** * Added a new release-notes entry covering gateway recovery, authentication, network policy/provenance output, uninstall safety, Windows bootstrap diagnostics, messaging defaults, and inference setup guidance. * Clarified Windows preparation steps around reboot behavior and redacting troubleshooting transcripts. * Expanded command reference details for OpenClaw wrapper behavior and new auto-pair tuning options. * Improved troubleshooting guidance for plugin installation issues, WSL repair/reboot cases, and install timing problems.  Signed-off-by: Julie Yaunches <jyaunches@nvidia.com> Co-authored-by: Prekshi Vyas <34834085+prekshivyas@users.noreply.github.com>

Dongni-Yang added 3 commits June 24, 2026 09:39

coderabbitai Bot reviewed Jun 26, 2026

View reviewed changes

Comment thread src/lib/validation.ts Outdated

Dongni-Yang added 3 commits June 26, 2026 14:45

merge(onboard): sync with upstream/main to resolve test-file-size-bud…

7688dc2

…get divergence Refs #4127 Signed-off-by: Dongni Yang <dongniy@nvidia.com>

Dongni-Yang assigned Dongni-Yang and unassigned Dongni-Yang Jun 26, 2026

Dongni-Yang added the v0.0.69 Release target label Jun 26, 2026

wscurran added area: onboarding Onboarding FSM, provider setup, sandbox launch, or first-run flow bug-fix PR fixes a bug or regression integration: openclaw OpenClaw integration behavior labels Jun 26, 2026

ahunnargikar-nvidia requested a review from prekshivyas June 26, 2026 16:44

ahunnargikar-nvidia assigned prekshivyas Jun 26, 2026

cv added v0.0.71 Release target and removed v0.0.69 Release target labels Jun 28, 2026

prekshivyas approved these changes Jun 29, 2026

View reviewed changes

fix(onboard): correlate plugin install network errors

49db962

github-advanced-security AI found potential problems Jun 30, 2026

View reviewed changes

Comment thread src/lib/validation.ts Fixed

test(onboard): isolate plugin failure classification

4d9b8f6

Signed-off-by: Carlos Villela <cvillela@nvidia.com>

cv added 2 commits June 29, 2026 21:51

Merge remote-tracking branch 'origin/main' into fix/4127-plugin-insta…

85fa59c

…ll-network-policy-hint

fix(onboard): require plugin installer command

8d4ee7e

Signed-off-by: Carlos Villela <cvillela@nvidia.com>

cv merged commit 0690f16 into main Jun 30, 2026
47 checks passed

cv deleted the fix/4127-plugin-install-network-policy-hint branch June 30, 2026 06:19

jyaunches mentioned this pull request Jun 30, 2026

docs: refresh v0.0.71 release docs #6070

Merged

21 tasks

Uh oh!

Conversation

Dongni-Yang commented Jun 26, 2026 • edited by cv Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Changes

Source Boundary and Removal Condition

Type of Change

Quality Gates

Verification

Uh oh!

coderabbitai Bot commented Jun 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

Estimated code review effort

Possibly related issues

Suggested labels

Suggested reviewers

Poem

❌ Failed checks (1 warning)

Uh oh!

github-code-quality Bot commented Jun 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Code Coverage Overview

TypeScript / code-coverage/plugin

TypeScript / code-coverage/cli

Uh oh!

github-actions Bot commented Jun 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

PR Review Advisor — No blocking findings

Uh oh!

github-actions Bot commented Jun 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

E2E Advisor Recommendation

E2E Recommendation Advisor

Required E2E

Optional E2E

New E2E recommendations

Dispatch hint

Uh oh!

github-actions Bot commented Jun 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Vitest E2E Scenario Recommendation

Vitest E2E Scenario Advisor

Required Vitest E2E scenarios

Optional Vitest E2E scenarios

Relevant changed files

Uh oh!

coderabbitai Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Dongni-Yang commented Jun 26, 2026

Uh oh!

github-actions Bot commented Jun 26, 2026

Selective E2E Results — ✅ All requested jobs passed

Uh oh!

prekshivyas left a comment

Choose a reason for hiding this comment

Uh oh!

github-actions Bot commented Jun 30, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

E2E Target Recommendation

E2E Target Advisor

Required E2E targets

Optional E2E targets

Relevant changed files

Uh oh!

Uh oh!

github-actions Bot commented Jun 30, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

PR Review Advisor (Nemotron Ultra) — No blocking findings

Action checklist

Findings index

⚠️ Resolve or justify before merge

💡 In-scope improvements

PRA-1 Improvement — Test uses hardcoded plugin version that will age

PRA-2 Improvement — Hint references single feature flag as example; consider generalizing if more plugin-dependent features exist

PRA-1 Improvement — Test uses hardcoded plugin version that will age

Dongni-Yang commented Jun 26, 2026 •

edited by cv

Loading

coderabbitai Bot commented Jun 26, 2026 •

edited

Loading

github-code-quality Bot commented Jun 26, 2026 •

edited

Loading

github-actions Bot commented Jun 26, 2026 •

edited

Loading

github-actions Bot commented Jun 26, 2026 •

edited

Loading

github-actions Bot commented Jun 26, 2026 •

edited

Loading

github-actions Bot commented Jun 30, 2026 •

edited

Loading

github-actions Bot commented Jun 30, 2026 •

edited

Loading

`PRA-1` Improvement — Test uses hardcoded plugin version that will age

`PRA-2` Improvement — Hint references single feature flag as example; consider generalizing if more plugin-dependent features exist

`PRA-1` Improvement — Test uses hardcoded plugin version that will age

`PRA-2` Improvement — Hint references single feature flag as example; consider generalizing if more plugin-dependent features exist