fix(ci): trigger build and review workflows for bot-created PRs

[MichaelFisher1997]

Summary

Bot-created PRs (from opencode/schedule-* or opencode/dispatch-* branches) never trigger the required build, unit-test, and ai-merge-gate status checks because GITHUB_TOKEN-created PRs don't fire pull_request events.

This leaves the dev ruleset permanently stuck at "Expected — Waiting for status to be reported" (see #443).

Changes

build.yml

• Add workflow_dispatch trigger with ref input for on-demand runs
• Short-circuit the changes filter job for dispatch events (test PRs always have code changes)
• Pass ${{ inputs.ref }} to all actions/checkout steps (build, unit-test, integration-test)

opencode-test-writer.yml

• Fix PR search: use --label automated-test instead of broken --head "opencode/dispatch-" filter
• Fetch actual branch name from the PR and pass it via --field ref="$BRANCH" to build.yml

Flow

test-writer → creates PR → trigger step
  ├→ gh workflow run build.yml --field ref="opencode/schedule-..."  → build + unit-test
  └→ gh workflow run opencode-pr.yml --field pr_number=N            → AI review + ai-merge-gate

All 3 required checks (build, unit-test, ai-merge-gate) will now post against the bot PR's head SHA.

Testing

• workflow_dispatch with GITHUB_TOKEN is the one exception that CAN trigger other workflows (no PAT needed)
• Human PRs are unaffected — they still fire pull_request events normally
• No duplicate runs: bot PRs never trigger pull_request, and dispatch concurrency is keyed separately

[github-actions]

📋 Summary

PR #446 adds workflow_dispatch trigger support to build.yml and fixes the PR search logic in opencode-test-writer.yml so bot-created PRs properly trigger required CI checks. This addresses issue #443 where GITHUB_TOKEN-created PRs don't fire pull_request events, leaving the dev ruleset stuck waiting for status checks.

---

📌 Review Metadata

• Reviewed Commit SHA: a39f6d3f7e1fa62aaab8d2584f973673d1943ea5
• Reviewed PR: fix(ci): trigger build and review workflows for bot-created PRs #446
• Linked Issue: References Added 33 tests across 3 new test files for world/worldgen modules. #443 (dev ruleset stuck waiting for bot PR status checks)
• Issue Verification: The PR fully addresses Added 33 tests across 3 new test files for world/worldgen modules. #443 by providing workflow_dispatch as an alternative trigger mechanism

---

🔴 Critical Issues (Must Fix - Blocks Merge)

None identified — this is a CI/CD workflow fix with no application code changes.

---

⚠️ High Priority Issues (Should Fix)

None identified.

---

💡 Medium Priority Issues (Nice to Fix)

[MEDIUM] .github/workflows/opencode-test-writer.yml:233 - PR selection may race with concurrent test-writer runs
Confidence: Medium
Description: The query gh pr list --base dev --label automated-test --state open --json number --jq '.[0].number' selects the first PR by creation time. If multiple test-writer runs execute concurrently or if previous bot PRs remain open, this could select the wrong PR.
Impact: Wrong workflow triggered for incorrect PR, potentially causing merge gate to run against wrong PR or wrong branch being built.
Suggested Fix: Consider adding --sort created --order desc explicitly or filtering by head branch pattern if multiple concurrent runs are possible.

---

ℹ️ Low Priority Suggestions (Optional)

[LOW] .github/workflows/build.yml:10 - workflow_dispatch ref input has no default
Description: The ref input is marked required: true but lacks a default. If the workflow is manually dispatched without providing ref, it will fail.
Suggested Fix: Add default: ${{ github.ref }} to provide the current ref as a sensible default for manual runs.

---

📊 SOLID Principles Score

Not applicable — this PR modifies GitHub Actions workflows, not Zig application code. SOLID principles apply to software architecture, not CI/CD configuration.

Principle	Score	Notes
Single Responsibility	N/A	Workflow configuration
Open/Closed	N/A
Liskov Substitution	N/A
Interface Segregation	N/A
Dependency Inversion	N/A
Average	N/A

---

🎯 Final Assessment

Overall Confidence Score: 85%

How to interpret: High confidence — the CI fix is sound and addresses the reported issue.

Confidence Breakdown:

• Code Quality: 90% (clean workflow syntax, proper use of GitHub Actions patterns)
• Completeness: 85% (addresses the core issue; minor edge cases around concurrent runs)
• Risk Level: 10% (low risk — workflow-only changes, doesn't affect runtime)
• Test Coverage: N/A (CI workflows not tested via zig build test)

Merge Readiness:

• All critical issues resolved
• SOLID average score >= 6.0 (N/A but trivially satisfied)
• Overall confidence >= 60%
• No security concerns
• Tests present and passing (N/A for workflow changes)

Verdict:

MERGE — The PR correctly fixes the bot PR CI trigger issue by adding workflow_dispatch support and properly passing branch refs to checkout steps. The changes are minimal, focused, and solve the stated problem.

---

Machine Readable Verdict

{
  "reviewed_sha": "a39f6d3f7e1fa62aaab8d2584f973673d1943ea5",
  "critical_issues": 0,
  "high_priority_issues": 0,
  "medium_priority_issues": 1,
  "overall_confidence_score": 85,
  "recommendation": "MERGE"
}

opencode session  |  github run

[github-actions]

📋 Summary

PR #446 adds workflow_dispatch trigger support to build.yml and fixes PR search logic in opencode-test-writer.yml so bot-created PRs properly trigger required CI checks. This fully addresses issue #443 where GITHUB_TOKEN-created PRs don't fire pull_request events.

---

📌 Review Metadata

• Reviewed Commit SHA: 899c472eb9b782e2397ad2e1fd6b250475a4be81
• Reviewed PR: fix(ci): trigger build and review workflows for bot-created PRs #446
• Linked Issue: References Added 33 tests across 3 new test files for world/worldgen modules. #443 (dev ruleset stuck waiting for bot PR status checks)
• Issue Verification: The PR fully addresses Added 33 tests across 3 new test files for world/worldgen modules. #443 via workflow_dispatch as an alternative trigger

---

🔴 Critical Issues (Must Fix - Blocks Merge)

None identified.

---

⚠️ High Priority Issues (Should Fix)

None identified.

---

💡 Medium Priority Issues (Nice to Fix)

✅ [FIXED] Previous issue: PR selection may race with concurrent test-writer runs
The fix in commit 899c472 now uses BRANCH="${{ steps.select-module.outputs.branch }}" and explicitly fetches the actual branch via gh api ... --jq '.head.ref', then passes it to build.yml --field ref="$BRANCH". This eliminates the race condition where the wrong PR could be selected.

---

ℹ️ Low Priority Suggestions (Optional)

[LOW] .github/workflows/build.yml:10 - workflow_dispatch ref input has no default
Description: The ref input is marked required: true but lacks a default. If the workflow is manually dispatched without providing ref, it will fail.
Suggested Fix: Add default: ${{ github.ref }} to provide the current ref as a sensible default for manual runs.

---

📊 SOLID Principles Score

Not applicable — this PR modifies GitHub Actions workflows, not Zig application code.

Principle	Score	Notes
Single Responsibility	N/A	Workflow configuration
Open/Closed	N/A
Liskov Substitution	N/A
Interface Segregation	N/A
Dependency Inversion	N/A
Average	N/A

---

🎯 Final Assessment

Overall Confidence Score: 92%

How to interpret: High confidence — the CI fix is sound and addresses the reported issue.

Confidence Breakdown:

• Code Quality: 92% (clean workflow syntax, proper concurrency grouping, correct use of GitHub Actions patterns)
• Completeness: 90% (addresses core issue; the race condition fix in commit 899c472 resolves the medium priority concern)
• Risk Level: 8% (very low risk — workflow-only changes, doesn't affect runtime)
• Test Coverage: N/A (CI workflows not tested via zig build test)

Merge Readiness:

• All critical issues resolved
• SOLID average score >= 6.0 (N/A but trivially satisfied)
• Overall confidence >= 60%
• No security concerns
• Tests present and passing (N/A for workflow changes)

Verdict:

MERGE — The PR correctly fixes the bot PR CI trigger issue with the second commit adding the critical race condition fix. The changes are minimal, focused, and solve the stated problem.

---

Machine Readable Verdict

{
  "reviewed_sha": "899c472eb9b782e2397ad2e1fd6b250475a4be81",
  "critical_issues": 0,
  "high_priority_issues": 0,
  "medium_priority_issues": 0,
  "overall_confidence_score": 92,
  "recommendation": "MERGE"
}

opencode session  |  github run