fix(release): resolve version-qualified catalog overrides

[Yash-Singh1]

Problem

Release run 37438990331 failed in every macOS and Linux desktop build because the catalog resolver looked up undici@^8 instead of undici. Windows builds then failed waiting for the Linux CLI archives that those jobs never uploaded.

Change

Strip the version range from the final package in an override selector before looking up a bare catalog: spec. Preserve the original override key and scoped package names. The shared resolver fixes both desktop staging and CLI archive staging.

Add regression coverage for undici@^8, ws@^8, scoped packages, and parent-qualified selectors.

Scope and approval

This is a focused fix for an obvious release build failure. It corrects the existing catalog lookup without changing dependency versions or product behavior, so it qualifies for the small bug fix exception.

Verification

• The new regression test failed before the fix with the exact undici@^8 catalog lookup error from the release logs.
• vp test run scripts/lib/resolve-catalog.test.ts passed all 3 tests after the fix.
• Ran the resolver against the current pnpm-workspace.yaml; all 56 overrides resolved, including undici@^8 and ws@^8.
• Targeted lint, formatting, and git diff --check passed.
• The desktop build test suite could not load because the local dependency install is missing effect/process. A full release build was not run locally.

Model: GPT-6. Harness: Codex.

[macroscopeapp]

Approvability

Verdict: Approved at 4b3afd6

Macroscope's review found this PR approvable — This narrowly scoped release-build fix correctly resolves version-qualified catalog override selectors while preserving existing dependency keys and behavior. The added unit coverage exercises direct, scoped, and nested selectors, with no product-default or static-analysis changes.

^{You can add or adjust custom eligibility rules. Learn more.}

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

• Configuration used: Path: .coderabbit.config.ts
• Review profile: CHILL
• Plan: Team
• Run ID: ffc6b6d9-ec40-4904-a529-521191aa66af

📥 Commits

Reviewing files that changed from the base of the PR and between a7a2230 and 4b3afd6.

📒 Files selected for processing (2)

• scripts/lib/resolve-catalog.test.ts
• scripts/lib/resolve-catalog.ts

Included review availability: This review used your included allowance. Your plan provides up to 10 included reviews per hour; 9 remain after this review.

---

📝 Walkthrough

Walkthrough

The catalog resolver now derives package names from version-qualified selectors when a bare catalog: spec has no explicit key. Added tests cover scoped and unscoped packages, parent-qualified overrides, and preservation of the original selectors.

Changes

Catalog selector resolution

Layer / File(s)	Summary
Derive lookup key and verify selector results scripts/lib/resolve-catalog.ts, scripts/lib/resolve-catalog.test.ts	The resolver removes a version suffix from the final selector when deriving a package name. A nonempty explicit catalog key remains the lookup key. Tests check version-qualified selectors and parent-qualified overrides. The result retains the original selectors as keys.

Priority: ⬆️ High

Estimated code review effort: 2 (Simple) | ~10 minutes

Change: Bug fix

Suggested reviewers: juliusmarminge

Merge Risk: ⚪ Minimal · up to 4b3af

No actionable merge-blocking issue is identified; the change is mergeable after normal checks.

🚥 Pre-merge checks | ✅ 4

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly identifies the release fix: resolving version-qualified catalog overrides.
Description check	✅ Passed	The description covers the problem, change, scope rationale, and verification results. It also states the local test limitation and that a full release build was not run.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

📝 Generate docstrings

• Commit to this branch
• Create a new PR

🧪 Generate unit tests (beta)

• Commit to this branch
• Create a new PR

• Autopilot · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

---

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

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

ℹ️ The exact PR base did not have a successful artifact. Baseline uses the latest successful main measurement shown below.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	4.9 KiB	5.0 KiB	+23 B (+0.5%)	6.8 KiB	✅
Codex	Thread snapshot wire	3.8 KiB	3.8 KiB	0 B (0.0%)	4.9 KiB	✅
Codex	Live turn WebSocket wire	1.2 KiB	1.2 KiB	+23 B (+1.9%)	2.0 KiB	✅
Codex	Live turn WebSocket decoded	20.8 KiB	20.9 KiB	+41 B (+0.2%)	29.3 KiB	✅
Codex	Live turn messages	1	2	+1 (+100.0%)	8	✅
Claude	Total thread wire	5.0 KiB	5.0 KiB	−17 B (−0.3%)	6.8 KiB	✅
Claude	Thread snapshot wire	3.8 KiB	3.8 KiB	0 B (0.0%)	4.9 KiB	✅
Claude	Live turn WebSocket wire	1.2 KiB	1.2 KiB	−17 B (−1.4%)	2.0 KiB	✅
Claude	Live turn WebSocket decoded	21.2 KiB	21.2 KiB	0 B (0.0%)	29.3 KiB	✅
Claude	Live turn messages	2	2	0 (0.0%)	8	✅

Baseline: 250e052 · PR result: 4b3afd6 · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 108.5 KiB
• Claude decoded thread snapshot: 108.8 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.