fix: preserve reviewer results after tool failures

[zzwong]

Summary

Closes #550.

The pipeline assigns incomplete_tool after obtaining a reviewer result when the required diff tool failed, did not finish, or was not invoked. The shared coverageResultProduced helper omitted that status, so consumers incorrectly treated a returned result as missing: the findings table displayed “did not run,” JSON emitted ran: false, and coverage details omitted known-empty skipped-file and constraint lists.

This recognizes incomplete_tool as a produced result. Readers can see what the reviewer actually returned while the coverage warning explains why the review cannot approve.

Before and after

For a reviewer that returned two findings, inspected its assigned file, and reported no skips or constraints, but whose diff-tool evidence was unsuccessful:

Surface	Before	After
Findings table	⚠️ did not run	2
JSON reviewer row	"findings": 2, "ran": false	"findings": 2, "ran": true
Coverage details	Empty skipped/constraint fields omitted	skipped: none; constraints: none

The same distinction applies to a real zero-finding result: it displays 0. A reviewer that produced no result still displays “did not run.” The tool-failure warning remains visible, and incomplete coverage still clamps an otherwise approving review to comment.

Implementation and scope

The production change adds incomplete_tool to the existing result-produced classification shared by Markdown rendering and the JSON view. Regression coverage exercises those consumers with zero and nonzero counts, known-empty lists, and genuinely missing results. Existing tests continue to cover the warning and approval withholding.

This is a focused reporting correction. #584 separately documents the persisted tool-evidence contract.

Verification

• The new Markdown and JSON regressions failed with the previous classification and passed with the fix.
• All 143 tests in internal/reviewplan and internal/view passed. Formatting, lint, and build checks passed with Go 1.26.3.
• CI passed, including tests, the 1Password opt-out configuration, static smoke tests, lint, and Linux/macOS/Windows builds.

The local full-suite run encountered a separate validation limitation: TestPiRPCReviewerExtensionLoadsInInstalledPi timed out during the installed Pi help preflight, and a focused retry reproduced that timeout. The affected-package tests and the CI test jobs above passed.

[monit-reviewer]

Automated PR Review

Reviewed commit: 2228b4107641
Profile: open-cli-collective-review - Posting as: monit-reviewer

Summary

Reviewer	Findings
go:implementation-tests	0
structure:repo-health	0

Reviewer Coverage

• go:implementation-tests — complete (broad); skipped: none; constraints: Review limited to the assigned changed Go files and read-only repository inspection; tests were not executed.
• structure:repo-health — complete (broad); skipped: none; constraints: Review limited to the three assigned changed files and their immediate status-producer/JSON-renderer context.

Inspected files (3)

• internal/reviewplan/summary.go
• internal/reviewplan/summary_test.go
• internal/view/review_reviewer_ran_test.go

0 PR discussion threads considered. 0 summarized; 0 resolved.

---

Completed in 1m 46s | $0.04 | openai-codex/gpt-5.6-terra | cr 0.10.299

Field	Value
Model	openai-codex/gpt-5.6-terra
Reviewers	go:implementation-tests, structure:repo-health
Engine	pi_rpc · openai-codex/gpt-5.6-terra
Reviewed by	cr · monit-reviewer
Duration	1m 46s wall · 2m 47s compute
Cost	$0.04
Tokens	14.2k in / 617 out

Per-workstream usage

• orchestrator-selection — openai-codex/gpt-5.6-terra
  • In: 5.8k
  • Out: 178
  • Cache read: 0
  • Cache create: 0
  • Cost: $0.01
  • Duration: 5s
• go:implementation-tests — openai-codex/gpt-5.6-terra
  • In: 4.5k
  • Out: 203
  • Cache read: 20.0k
  • Cache create: 0
  • Cost: $0.02
  • Duration: 1m 27s
• structure:repo-health — openai-codex/gpt-5.6-terra
  • In: 2.6k
  • Out: 199
  • Cache read: 20.0k
  • Cache create: 0
  • Cost: $0.01
  • Duration: 1m 11s
• orchestrator-rollup — openai-codex/gpt-5.6-terra
  • In: 1.2k
  • Out: 37
  • Cache read: 2.8k
  • Cache create: 0
  • Cost: $0.00
  • Duration: 2s