fix(installer): fail closed when a present OpenShell cannot report its version (#7300) - #7305
Conversation
…s version (#7300) When `openshell` exists on PATH but exits non-zero (cannot report a version), `maybe_install_openshell_during_install if-missing` (source-checkout branch) returned 0 on `command_exists openshell` alone — an existence check with no health check. So the installer silently skipped the `Installing OpenShell CLI` step, and the preinstall version gate that would catch it (preinstall_backup_and_retire_legacy_gateway, added in #7078) only runs when a sandbox is registered — on a host with no sandbox it early-returns. The result was onboarding proceeding to exit 0 against an undeterminable OpenShell version. Health-check the if-missing guard: when openshell is present but `installed_openshell_version` is empty, `error` out (non-zero exit) with a clear message, leaving gateway and sandbox state unchanged, instead of keeping a broken binary. A binary that reports a version is still preserved (the #3989 developer-autonomy intent); an absent binary still triggers the install. Verified in real bash on Ubuntu 24.04: sourcing scripts/install.sh with a stub `openshell` (exit 1) makes `maybe_install_openshell_during_install if-missing` abort with the new error; a stub that reports 0.0.85 is kept with exit 0. Two regression tests pin both. Signed-off-by: Yanyun Liao <yanyunl@nvidia.com>
|
Note Reviews pausedIt looks like this branch is under active development. To avoid overwhelming you with review comments due to an influx of new commits, CodeRabbit has automatically paused this review. You can configure this behavior by changing the Use the following commands to manage reviews:
Use the checkboxes below for quick actions:
📝 WalkthroughWalkthroughThe installer now fails closed when an existing OpenShell binary cannot report its version. Tests cover invalid binaries, absent binaries, deferred installation, and supported versions. ChangesOpenShell version validation
Estimated code review effort: 2 (Simple) | ~10 minutes Suggested labels: 🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Comment |
Code Coverage OverviewLanguages: TypeScript TypeScript / code-coverage/pluginThe overall coverage in commit d73548d in the TypeScript / code-coverage/cliThe overall coverage in commit d73548d in the Show a code coverage summary of the most impacted files.
Updated |
PR Review Advisor — InformationalAdvisor assessment: Informational / high confidence Model lanes
Nemotron output stays in workflow artifacts and does not change the assessment above. E2E guidanceAdvisory only. E2E / PR Gate selects and runs jobs independently. Recommended E2E: 1 optional E2E recommendation
This automated review informs maintainers. Warnings and suggestions do not require a response. A maintainer decides whether to merge. |
…le gate (#7300) Address PR Review Advisor PRA-1 (blocker): the previous check lived inside the if-missing branch, after the NEMOCLAW_DEFER_OPENSHELL_INSTALL early return, so a deferred install (env set) combined with a no-sandbox run (preinstall version gate skipped at install.sh:2276) could still proceed against an unreportable OpenShell version. Move it to require_reportable_openshell_version(), called unconditionally in main() after install_nemoclaw/verify_nemoclaw and before onboarding — independent of defer, sandbox count, and if-missing/force mode. The force path has already reinstalled openshell (healthy) by then; a present-but-unreportable binary on the source-checkout path fails closed; an absent binary or one that reports a version passes. Tests now drive the gate directly, including a defer-set case proving it is not bypassable. Verified live on Ubuntu 24.04: sourcing scripts/install.sh and calling the gate with a stub openshell (exit 1) aborts with the diagnostic even under NEMOCLAW_DEFER_OPENSHELL_INSTALL=1; a stub reporting 0.0.85 passes. Signed-off-by: Yanyun Liao <yanyunl@nvidia.com>
There was a problem hiding this comment.
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 `@scripts/install.sh`:
- Line 4012: The installer currently mutates gateway state before validating an
existing openshell binary. In scripts/install.sh:4012-4012, invoke
require_reportable_openshell_version before
preinstall_backup_and_retire_legacy_gateway while preserving the absent-binary
installation path; in test/install-openshell-upgrade-prompt.test.ts:216-263, add
an end-to-end flow test confirming a broken existing binary is rejected without
modifying gateway or sandbox state.
🪄 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: be648129-006a-4ef1-b3f9-e4bdcbb058e0
📒 Files selected for processing (2)
scripts/install.shtest/install-openshell-upgrade-prompt.test.ts
|
✨ Related open issues: |
…enshell-unreportable-version
Run the version gate before legacy gateway backup and retirement. Keep the post-install gate to verify the installed binary. Test the mutation boundary and absent-binary path through the installer main flow. Signed-off-by: Julie Yaunches <jyaunches@nvidia.com>
There was a problem hiding this comment.
🧹 Nitpick comments (1)
test/install-openshell-upgrade-prompt.test.ts (1)
308-324: 📐 Maintainability & Code Quality | 🔵 Trivial | ⚡ Quick winExercise the healthy existing-binary path through
main().The healthy case only tests the gate directly; it does not prove a reportable existing
openshellsurvives the installer flow. Add arunInstallerOpenshellVersionFlow(healthyOpenshell)case that asserts success and that the original binary remains unchanged. Based on PR objectives, healthy existing binaries must remain preserved. As per path instructions, prefer observable outcomes through the public boundary.🤖 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 `@test/install-openshell-upgrade-prompt.test.ts` around lines 308 - 324, Add a test in the existing installer flow suite using runInstallerOpenshellVersionFlow(healthyOpenshell) to exercise main() with a reportable binary already present. Assert successful completion and verify the original healthy openshell binary remains unchanged, using observable installer outputs rather than testing the gate directly.Source: Path instructions
🤖 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 `@test/install-openshell-upgrade-prompt.test.ts`:
- Around line 308-324: Add a test in the existing installer flow suite using
runInstallerOpenshellVersionFlow(healthyOpenshell) to exercise main() with a
reportable binary already present. Assert successful completion and verify the
original healthy openshell binary remains unchanged, using observable installer
outputs rather than testing the gate directly.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Enterprise
Run ID: fdc52c04-ce44-49d7-bc48-472b7c6e86b3
📒 Files selected for processing (2)
scripts/install.shtest/install-openshell-upgrade-prompt.test.ts
Pass fixture setup as a callback so the changed test file adds no if statement. Preserve the broken-binary and absent-binary installer flow cases. Signed-off-by: Julie Yaunches <jyaunches@nvidia.com>
Reject version text when the OpenShell command exits unsuccessfully. Stop before gateway or sandbox mutation and cover both failed and healthy commands. Signed-off-by: Julie Yaunches <jyaunches@nvidia.com>
Prove the second gate rejects a failed version command produced by installation before onboarding. Document the version helper's nonzero return contract. Signed-off-by: Julie Yaunches <jyaunches@nvidia.com>
Provide the Station vLLM helper expected after SCRIPT_DIR is redirected in the installer-flow test. Signed-off-by: Julie Yaunches <jyaunches@nvidia.com>
Signed-off-by: Julie Yaunches <jyaunches@nvidia.com>
Signed-off-by: Aaron Erickson <aerickson@nvidia.com>
Signed-off-by: Aaron Erickson <aerickson@nvidia.com>
Signed-off-by: Aaron Erickson <aerickson@nvidia.com>
Signed-off-by: Aaron Erickson <aerickson@nvidia.com>
|
🌿 Preview your docs: https://nvidia-preview-pr-7305.docs.buildwithfern.com/nemoclaw |
Signed-off-by: Aaron Erickson <aerickson@nvidia.com>
Signed-off-by: Aaron Erickson <aerickson@nvidia.com>
ericksoa
left a comment
There was a problem hiding this comment.
Reviewed exact head d73548d on current main 9171d79. Registered-sandbox installs now complete the required strict backup before failing closed on an unreportable OpenShell, without gateway retirement or registry mutation; the post-install gate protects deferred and no-registry paths. Focused installer tests, full exact-head CI, credentialed cloud-onboard E2E, final docs receipt, both advisor lanes, CodeRabbit, DCO/signatures, resolved threads, and the security sweep are green.
<!-- markdownlint-disable MD041 --> ## Summary Add the canonical `docs/changelog/2026-07-25.mdx` release entry with the exact `## v0.0.96` heading. The entry reconciles all 90 first-parent commits since v0.0.95 with all 92 merged PRs in the live `v0.0.96` label ledger and groups the user-visible changes by operator journey. ## Changes - Add the parser-safe dated MDX changelog entry for v0.0.96 with root-absolute links to the focused user guides. - Source summary: - [#7194](#7194) -> `docs/changelog/2026-07-25.mdx`: Document persistent baseline network policy exclusions and their inspection, rebuild, and snapshot behavior. - [#7188](#7188), [#7427](#7427), and [#7546](#7546) -> `docs/changelog/2026-07-25.mdx`: Document DNS-backed HTTPS inference routing, keyless loopback endpoints, and provider-marker isolation. - [#7238](#7238) -> `docs/changelog/2026-07-25.mdx`: Document blueprint sandbox and provider identifier validation before state writes or OpenShell calls, with bounded terminal-safe rejection previews. - [#7319](#7319), [#7274](#7274), [#7528](#7528), [#7353](#7353), and [#7560](#7560) -> `docs/changelog/2026-07-25.mdx`: Document the managed default gateway service, onboarding readiness, and container-runtime identity safeguards. - [#7349](#7349), [#7498](#7498), [#7406](#7406), [#7196](#7196), [#7559](#7559), [#7421](#7421), [#7510](#7510), [#7295](#7295), and [#7565](#7565) -> `docs/changelog/2026-07-25.mdx`: Document gateway-scoped status, lifecycle diagnostics, managed MCP recovery, delete-edge safeguards, and fail-closed CLI prompt and command output. - [#7591](#7591) -> `docs/changelog/2026-07-25.mdx`: Document opt-in authenticated MCP tool-name discovery, its bounded and names-only contract, probe interaction, and rebuild requirement. - [#7305](#7305), [#7480](#7480), [#7471](#7471), [#7365](#7365), and [#7541](#7541) -> `docs/changelog/2026-07-25.mdx`: Document installer version checks, version-tag reporting, license guidance, WSL Ollama selection, and DGX Station vLLM detection. - [#7482](#7482), [#7466](#7466), [#7208](#7208), [#7434](#7434), and [#7586](#7586) -> `docs/changelog/2026-07-25.mdx`: Document Ollama resource details, reasoning precedence, Hermes onboarding behavior, and preserved managed Hermes BuildKit failures. - [#6830](#6830), [#7492](#7492), [#7563](#7563), and [#7582](#7582) -> `docs/changelog/2026-07-25.mdx`: Document the authoritative OpenClaw production lock, fixed managed-image dependencies, immutable Hermes base adoption, and Hermes image-size reduction. - [#7505](#7505), [#7530](#7530), [#7547](#7547), [#7508](#7508), [#7548](#7548), [#7549](#7549), [#7537](#7537), [#7534](#7534), [#7515](#7515), [#7511](#7511), [#7551](#7551), [#7562](#7562), [#7575](#7575), [#7496](#7496), [#7594](#7594), [#7595](#7595), and [#7599](#7599) -> `docs/changelog/2026-07-25.mdx`: Summarize release validation, transient and bounded dispatch reconciliation, exact pre-tag qualification, identity revalidation, npm-audit retry, sharding, image reuse, timeout, telemetry, and workflow-hardening changes. - Reconciled without separate changelog prose: - [#7539](#7539), [#7526](#7526), [#7507](#7507), [#7506](#7506), [#7519](#7519), [#7516](#7516), [#7396](#7396), [#7254](#7254), [#7583](#7583), [#7596](#7596), and [#7598](#7598): Test-harness or fixture-only changes. - [#7403](#7403), [#7161](#7161), [#6877](#6877), [#7531](#7531), [#7525](#7525), [#7522](#7522), [#7536](#7536), [#7552](#7552), [#7566](#7566), [#7553](#7553), [#7561](#7561), [#7577](#7577), [#7569](#7569), [#7585](#7585), [#7584](#7584), [#7592](#7592), [#7580](#7580), [#7571](#7571), [#7517](#7517), [#7589](#7589), [#7402](#7402), [#7558](#7558), [#7544](#7544), and [#7601](#7601): Dependency, internal recovery, validation, contributor-workflow, E2E optimization, telemetry, or CI trust changes with no separate user-facing release claim. - [#7556](#7556), [#7573](#7573), [#7576](#7576), and [#7578](#7578): Experimental repository-maintainer conflict automation with no canonical user documentation surface. ## Type of Change - [ ] Code change (feature, bug fix, or refactor) - [ ] Code change with doc updates - [x] Doc only (prose changes, no code sample modifications) - [ ] Doc only (includes code sample changes) ## Quality Gates - [ ] Tests added or updated for changed behavior - [x] Existing tests cover changed behavior — justification: `test/changelog-docs.test.ts` validates dated changelog structure, version headings, and published links. - [ ] Tests not applicable — justification: - [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: ## Documentation Writer Review - [x] Documentation writer subagent reviewed the completed changes - Result: `docs-updated` - Evidence: Reviewed `docs/changelog/2026-07-25.mdx` at exact head `0f5dedb47` against 90 first-parent release commits and 92 merged PRs labeled `v0.0.96`. Verified parser-safe MDX SPDX, the exact version heading, literal CLI names, writing style, skip terms, all 20 root-absolute published links, and the accepted #7591 opt-in authenticated discovery bounds. #7544, #7599, and #7601 remain internal or CI-only release-ledger entries. Changelog tests passed 6/6, the docs build passed with 0 errors and two pre-existing Fern warnings, and `npm run check:diff` plus the final diff check passed. - Agent: Codex Desktop documentation-writer subagent <!-- docs-review-head-sha: 0f5dedb --> <!-- docs-review-agents-blob-sha: be20a09 --> ## DGX Station Hardware Evidence - [ ] Tested on DGX Station - Tested commit: - Station profile/scenario: - Result: - Supporting evidence: ## Verification - [x] PR description includes a `Signed-off-by:` line and every commit appears as `Verified` in GitHub - [x] Normal `pre-commit`, `commit-msg`, and `pre-push` hooks passed, or `npm run check:diff` passed when hooks were skipped or unavailable - [x] Targeted behavior tests pass for the current change set, or tests are marked not applicable above — `npx vitest run test/changelog-docs.test.ts`: 6/6 passed. - [ ] Applicable broad gate passed — `npm test` for broad runtime/test-harness changes; `npm run check` for repo-wide validation/coverage changes — command/result: Not applicable to this prose-only changelog entry. - [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) — the build passed with 0 errors and 2 existing Fern warnings; the published-route check passed. - [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) — native changelog files use the required parser-safe MDX SPDX comment and no frontmatter. --- Signed-off-by: Carlos Villela <cvillela@nvidia.com> <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Persistent network policy exclusions with consistent restore/exclusion reporting across rebuilds/snapshots. * Opt-in MCP tool discovery via `mcp status --tools` with bounded, redacted authenticated traffic. * Improved HTTPS inference switching for custom endpoints and refreshed onboarding/model menu details. * Refined OpenShell gateway defaults for port `8080`, including more reliable readiness checks. * **Bug Fixes** * Prevent incorrect provider/model restoration after compatible-provider update failures. * Preserve managed MCP state after exec loss and tighten gateway/doctor status scoping. * **Tests** * Stronger, fail-closed release validation with hardened evidence/artifact handoff and bounded timeouts/retries. <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Signed-off-by: Prekshi Vyas <prekshiv@nvidia.com> Co-authored-by: Prekshi Vyas <prekshiv@nvidia.com>
Summary
When the installed
openshellbinary exists but exits non-zero (cannot report its version), the installer silently skipped the OpenShell version check and the[INFO] Installing OpenShell CLIstep, then completed onboarding and exited 0 — no error about the undeterminable version.Root cause
Two paths combine:
maybe_install_openshell_during_install if-missing(source-checkout branch,scripts/install.sh:1276) returns 0 oncommand_exists openshellalone — an existence check, not a health check. A present-but-broken stub is treated as a developer's own binary to keep, so theInstalling OpenShell CLIinstall is skipped.preinstall_backup_and_retire_legacy_gateway, the[ -n "$old_openshell_version" ] || error …atinstall.sh:2323, added in fix(installer): retire out-of-range gateway after backup #7078) only runs when a sandbox is registered — it early-returns atinstall.sh:2276whensandbox_countis 0.So on a host with no registered sandbox, neither gate fires and onboarding proceeds to exit 0 against an OpenShell whose version can't be determined. (With a sandbox, #7078's gate already aborts — the reporter's exit-0 transcript is pre-#7078 behavior.)
Fix
Health-check the
if-missingguard: whenopenshellis present butinstalled_openshell_versionis empty,errorout (non-zero exit,[ERROR]) with a clear, actionable message, leaving gateway and sandbox state unchanged — instead of silently keeping a broken binary. A binary that does report a version is still preserved (the #3989 developer-autonomy intent); an absent binary still triggers the normal install.Testing
test/install-openshell-upgrade-prompt.test.ts, sources the realscripts/install.sh): a presentopenshellthat exits non-zero →maybe_install_openshell_during_install if-missingfails closed with "could not report its version"; a presentopenshellthat reports0.0.85→ kept, exit 0.scripts/install.shwith a stubopenshell(exit 1) — before the fixmaybe_install_openshell_during_install if-missingreturned 0 with no install and no error; after the fix it aborts with the new[ERROR]. A stub reporting a version is kept with exit 0.Closes #7300
Signed-off-by: Yanyun Liao yanyunl@nvidia.com
Summary by CodeRabbit
Documentation Writer Review
docs-updateddocs/get-started/quickstart.mdxanddocs/manage-sandboxes/update-sandboxes.mdxnow describe version reportability and range validation after backup, matching the accepted Expected Result.docs/get-started/quickstart.mdx,docs/manage-sandboxes/update-sandboxes.mdx