Repository navigation
chore(goose): prepare verified v1.54.0 engine upgrade - #69137
Conversation
Update the Goose pin, archive checksum, smoke expectations, and runtime regression coverage. Keep validation gaps explicit in the draft PR. Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
|
✅ Design Decision Gate 🏗️ completed the design decision gate check. See the comment below for the result and any generated ADR draft. Design Decision Gate 🏗️Outcome: The agent completed and reported no action. A no-op output alone does not verify that all required work was completed or that publication was possible. Agent messages are omitted from this comment. The category uses structured run status, not interpretation of agent prose.
|
|
🧠 Matt Pocock Skills Reviewer has completed the skills-based review. ✅
|
|
✅ Test Quality Sentinel completed test quality analysis. Test Quality SentinelOutcome: The agent completed and reported no action. A no-op output alone does not verify that all required work was completed or that publication was possible. Agent messages are omitted from this comment. The category uses structured run status, not interpretation of agent prose.
|
|
✅ PR Code Quality Reviewer completed the code quality review.
|
|
✅ Ponytail Reviewer completed successfully! Ponytail ReviewerOutcome: The agent completed and reported no action. A no-op output alone does not verify that all required work was completed or that publication was possible. Agent messages are omitted from this comment. The category uses structured run status, not interpretation of agent prose.
|
![github-actions[bot]](https://avatars.githubusercontent.com/in/15368?s=60&v=4){kind=small}
There was a problem hiding this comment.
Verdict
I didn't find a merge-blocking issue in the changed lines.
What I checked
Version pin/checksum updates, the runtime harness environment isolation, the MCP tool allowlist regression, and the smoke/version fixtures all stay internally consistent with the Goose 1.54.0 bump.
🔎 Code quality review by PR Code Quality Reviewer · copilot · gpt54 · 20.4 AIC · ⌖ 5.49 AIC · ⊞ 21.4K
Comment /review to run again
There was a problem hiding this comment.
🔵 Needs a closer look
Current-source and native-runtime validation remain incomplete, and the generated-lock gate remains unresolved.
4 open findings
What changed in this PR
Updates gh-aw’s unsupported Goose sample to v1.54.0 and strengthens its integration regressions.
Changes:
- Adds the verified release checksum while retaining v1.53.0 overrides.
- Updates version expectations and smoke fixtures.
- Isolates the native runtime test and adds MCP allowlist checks.
| File | Description |
|---|---|
pkg/workflow/goose_workflow_test.go |
Updates execution expectations and CLI fixture versions. |
pkg/workflow/goose_smoke_test.go |
Updates smoke evidence and runtime probe fixtures. |
pkg/workflow/goose_runtime_test.go |
Isolates the runtime environment and checks tool allowlists. |
.github/workflows/smoke-goose.md |
Requires v1.54.0 verification evidence. |
.github/workflows/shared/goose.md |
Updates the default pin, checksum, and release link. |
.changeset/patch-goose-1-54.md |
Documents the upgrade and regression coverage. |
🧠 Review effort: Max
There was a problem hiding this comment.
Skills-Based Review 🧠
Applied /improve-codebase-architecture and /tdd (per pr-triage) to the Goose v1.53.0→v1.54.0 upgrade. All TestGoose* unit tests pass locally against the committed state; the mechanical version bump is correct (new checksum keyed by version, old 1.53.0 entry retained, fixtures updated consistently).
📋 Key Themes & Highlights
Key Themes
- Credential isolation improvement: goose_runtime_test.go now builds cmd.Env from an explicit allowlist (PATH/HOME/TMPDIR) instead of inheriting os.Environ(), matching the PR's stated goal of a synthetic-only test environment.
- New allowlist regression: an excluded MCP tool plus assertions in the handler and completion stub verify extension tool-leasing preserves the configured allowlist — good defense-in-depth, though this path is only exercised when GH_AW_TEST_GOOSE_BINARY is set, which is never the case in CI per the PR's own validation notes.
- Honest PR description: the author explicitly flags the stale lock file, unrun make fmt/make recompile, and unresolved live-authorization gate as outstanding — this draft should not be merged as-is, consistent with its title/labeling.
Positive Highlights
- ✅ Version bump is mechanically consistent across .github/workflows/shared/goose.md, smoke-goose.md, and all three pkg/workflow/goose_*_test.go files — no stray 1.53.0 references were missed in source (only the intentional backward-compat checksum entry and one unrelated pre-existing-version test case remain).
- ✅ gofmt -l is clean on the changed Go files.
- ✅ Checksum-before-extraction behavior is preserved; no new credential recipients.
No blocking issues found in the diff itself. Two inline suggestions posted to strengthen allowlist test coverage without requiring the opt-in native binary.
🧠 Reviewed using Matt Pocock's skills by Matt Pocock Skills Reviewer · copilot · sonnet50 · 124.8 AIC · ⌖ 14.6 AIC · ⊞ 11.1K
Comment /matt to run again
Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
Regenerate the four Goose workflow locks from merged sources, preserving the Goose v1.54.0 upgrade and upstream firewall, gateway, and artifact changes. Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
Check excluded MCP tool advertisements on every model turn, clarify opt-in native coverage, and cover tool filtering without a release binary. Preserve the latest base branch engine updates. Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
|
/smoke-goose |
|
🪿 smoke-goose has started processing this pull request comment |
|
🎉 This pull request is included in a new release. Release: |
Summary
Bring the unsupported Goose sample forward from v1.53.0 to the latest stable v1.54.0 release, published October 8, 2026. The release changes extension leasing/tool selection and usage accounting, so the upgrade keeps those integration boundaries explicit rather than changing shared engine defaults.
Pin the new release with its independently downloaded and SHA-256-verified Linux archive (
ab396c95aebbd28b4d38c175234be8d4a3ea0c96e7ccb24c307e4618c9be57e1), retain the existing verified v1.53.0 override, and update smoke/version fixtures. Extend the opt-in native runtime regression to check MCP tool allowlists and use a synthetic-only environment with disposable HOME/TMPDIR instead of inheriting ambient credentials.Static review of the tagged CLI/provider sources found the existing headless flags, OpenAI-compatible endpoint routing, native MCP configuration and JSONL contract remain supported. The existing unified-session parser already handles streaming text, correlated tools/failures, cumulative usage and cache-write normalization; no shared parser/schema changes are included.
Regenerate normal locks for all four consumers of
shared/goose.md: smoke, engine conformance, daily team-evolution insights, and daily token-consumption report. The generated changes are limited to the Goose version/checksum and corresponding metadata; no daily-workflow authored source, permissions or shared defaults changed.Validation and outstanding work
make recompilein a disposable validation clone, compiling 344/344 workflows (382 warnings reported).make fmtwith the repository's locked formatter dependencies restored in the validation clone.go test ./pkg/workflow ./pkg/cli -run '^TestGoose' -count=1 -v, including Goose compile/routing regressions and unified-session reconstruction. The opt-in native release runtime test was skipped; no release binary execution is claimed.npm run test:js -- parse_goose_log.test.cjs --no-file-parallelism(54 tests).PATH="/opt/homebrew/bin:$PATH" BASE_REF=da1f3ed6b199c8eb2eb3f7e67900abc768ed6753 make agent-report-progress-no-test, including Go/custom lint, schemas and repository-wide workflow drift validation (344/344 workflows; all lock files in sync). An earlier fullmake agent-report-progressran the unit sampler, but failed lint on two introduced issues; both were corrected before the successful final gate. Focused Goose tests were run separately because the sampler selected unrelated container-pin tests.git diff --checkand byte comparison of the five published files against the validated clone.4f57b32cd8275c6d6311c6417c90fcd712dbe02e(also run38823643332). Commit8fdba85aa847a06554276553968bafebe905f995supplies all four missing regenerated locks. No new hosted result is claimed.Source/control overview
Initial read-only security review found no confirmed vulnerabilities, but returned a blocked result for execution isolation/provenance and broad validation-hook coverage. Candidate mitigations and local compile/parser evidence are present; there is no revised reviewer PASS and no claim of live authorization.