Validate review-summary edge cases

[mchwang]

PR #9's final summary retained two moderate concerns and one styling nit despite reporting no findings. Validate all three before the real-issue experiment without inferring defects from summary shorthand.

Results:

• Manual reassignment after retry.ts → renamed.ts keeps both removed old-path and added new-path segments in scope.
• A literal declared filename *.txt is not confused with a decoy a.txt; planting rejects the history when the literal file disappears.
• Accepted changes render with the existing muted treatment and explicit Accepted outside plan label. A proposed error-color assertion failed because the summary supplied no expected alternative; no new color semantics were inferred.

No production defect reproduced, so this PR changes no runtime behavior. It adds focused real-Git validation fixtures for the two behavioral concerns and commits the evidence record required before issue #3's human experiment.

Validation on exact pushed head 2708ae9: typecheck, 190 unit/integration tests, and 39 browser tests passed locally; both required CI runs passed.

Review-lesson audit

• Originating PR Add the local plan-indexed review screen #9 summary-only concerns: handled by the existing Review readiness rule requiring reproduction or a concrete follow-up issue; issue Validate review-summary edge cases before the real-issue experiment #10 and this evidence PR are that follow-up.
• PR Validate review-summary edge cases #17 round 1, rename segment selection: fixed by asserting -/+ operations and resolving the removed side through oldPath and added side through path before reassignment. Captured by the new Review readiness rule requiring summary-concern fixtures to assert the disputed intermediate representation.
• PR Validate review-summary edge cases #17 round 1, requested literal-pathspec production fix: declined because direct Git controls prove ls-tree treats *.txt literally and rejects :(glob) magic as unsupported. The fixture now asserts those command semantics before the plant outcome. This command-specific false premise needs no production change and is covered by the same new fixture-evidence rule.
• PR Validate review-summary edge cases #17 round 2, claimed retained P3 fixture item: declined because the fixture replaces plan.items with a P1-only array. It now asserts the exact item IDs before import; typecheck and all focused assertions pass. This false intermediate-state premise is covered by the same new fixture-evidence rule.
• PR Validate review-summary edge cases #17 round 3: no findings. Its overview asked for the final in-scope assertion already added in the following commit.
• PR Validate review-summary edge cases #17 round 4, ambiguous validation commit wording: accepted. The evidence record now labels 5a2685c as the baseline and points exact-head results to this PR. A new rule requires evidence records to distinguish baselines, checkpoints, and validated heads.
• PR Validate review-summary edge cases #17 round 5 on exact head 2708ae9: no findings; approval recommended.
• All three review threads are resolved. The review loop converged after five rounds.

Deferred: issue #3's manual assignment and paired human go/no-go experiment remain pending.

Closes #10.

[Copilot]

Copilot review overview

🟡 Changes recommended

Two moderate validation issues and one documentation nit remain unresolved.

Get a fresh assessment by requesting another Copilot review.

Review effort: Lite
Findings: 2

Open (2)

• Fix planting check to use literal pathspec matching · New
• Select rename segments using both path and oldPath · New

What changed in this PR

Adds focused test fixtures and documentation for review-summary edge cases without intended runtime changes.

Changes:

• Tests rename reassignment behavior.
• Tests literal filename planting.
• Records styling validation and experiment status.

File	Summary	Review status
test/​review.test.ts	Rename reassignment fixture	Changes required: assert both path/oldPath and operations.
test/​plant.test.ts	Literal path planting fixture	Changes required: fix literal pathspec handling.
docs/​experiments/​review-summary-edge-cases.md	Validation evidence record	Nit: accurately describe rename segment fields.

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[Copilot]

Copilot review overview

🟡 Changes recommended

A critical fixture issue blocks the rename/reassignment assertions in test/review.test.ts.

Get a fresh assessment by requesting another Copilot review.

Review effort: Lite
Findings: 1

Open (1)

• Update P3 fixture after retry.ts is renamed · New

Resolved since last review (2)

• Select rename segments using both path and oldPath
• Fix planting check to use literal pathspec matching

[Copilot]

Copilot review overview

🔵 Needs a closer look

Add a final assertion confirming both renamed-file candidates remain in scope.

Review effort: Lite
Findings: None

Resolved since last review (1)

• Update P3 fixture after retry.ts is renamed

[Copilot]

Copilot review overview

🟢 Approval recommended

No blocking issues remain; the sole documentation nit is non-blocking.

Review effort: Lite
Findings: None

Previously missed (1)

In code that hasn't changed since last review

Evidence record references a different validation commit

docs/​experiments/​review-summary-edge-cases.md:3

The evidence record says it was validated from 5a2685c, but the PR metadata reports validation on the exact pushed head c4d580b. This makes the record point at a different tree than the claimed validation; update the recorded head (or explicitly explain why an earlier checkpoint is intentional).

[Copilot]

Copilot review overview

🟢 Approval recommended

All reviewed concerns are covered by passing focused validation and no runtime issues remain.

Review effort: Lite
Findings: None

[Copilot]

Copilot review overview

🟢 Approval recommended

All reviewed concerns are addressed, with no production behavior changes.

Review effort: Lite
Findings: None