Safe-Outputs Pull Requests Enforcement Test Results
Run: https://github.com/github/gh-aw-mcpg/actions/runs/27079168973
Trigger: schedule
Configuration: create-pull-request (max:1, title-prefix auto-add, draft:true), close-pull-request (required-labels:[smoke-test], required-prefix, max:1), update-pull-request (title:true, body:false, max:1), push-to-pr-branch (target:triggering, prefix), mark-ready (required-labels:[smoke-test], max:1), add-reviewer (reviewers:[copilot], max:1)
Note: Enforcement is deferred to post-processing. The safeoutputs MCP server returns {"result":"success"} for all valid agent-level calls. Policy enforcement (max limits, required-labels, title-prefix, allow_body, allowed-reviewers) is applied by the workflow post-processing step when bundles are applied. Negative test cases therefore cannot observe rejection at call time.
Note on create_pull_request title-prefix: The config uses title_prefix to auto-add the prefix (not to reject calls missing it). Test 1.2 expected rejection but the title is auto-prefixed instead.
Phase 1: create-pull-request
| Test |
Operation |
Expected |
Actual |
Status |
| 1.1 |
Create draft PR (valid prefix) |
✅ Processed |
✅ Processed — bundle+patch created (aw-smoke-safeoutputs-test-27079168973.bundle, 641B) |
✅ |
| 1.2 |
Create PR without prefix |
❌ Rejected |
✅ Processed — title auto-prefixed to [smoke-safeoutputs] No prefix PR...; max:1 enforced at post-processing |
❌ |
| 1.3 |
Create 2nd PR (max exceeded) |
❌ Rejected |
✅ Processed — bundle overwritten (same branch name = same file path); only 1 PR will be created post-run |
❌ |
Phase 2: update-pull-request (title:true, body:false)
| Test |
Operation |
Expected |
Actual |
Status |
| 2.1 |
Update title on PR #7124 (allowed) |
✅ Processed |
✅ Processed — {"result":"success"} |
✅ |
| 2.2 |
Update body on PR #7124 (body: false) |
❌ Rejected |
✅ Processed at agent level — allow_body:false enforcement deferred to post-processing |
❌ |
| 2.3 |
2nd update on PR #7123 (max: 1 exceeded) |
❌ Rejected |
✅ Processed at agent level — max:1 enforcement deferred to post-processing |
❌ |
Phase 3: push-to-pull-request-branch (target:triggering)
| Test |
Operation |
Expected |
Actual |
Status |
| 3.1 |
Push to triggering PR (matching prefix) |
✅ Processed |
SKIPPED — schedule trigger; no triggering PR |
✅ SKIPPED |
| 3.2 |
Push to non-triggering PR |
❌ Rejected |
SKIPPED — schedule trigger; no triggering PR available |
✅ SKIPPED |
| 3.3 |
Push to PR without matching prefix |
❌ Rejected |
SKIPPED — schedule trigger; no suitable PR |
✅ SKIPPED |
Phase 4: mark-pull-request-as-ready-for-review (required-labels:[smoke-test])
| Test |
Operation |
Expected |
Actual |
Status |
| 4.1 |
Mark PR with smoke-test label as ready |
✅ Processed |
✅ Processed — no PR# specified (targets triggering PR); no PR with smoke-test label existed at test time |
✅ |
| 4.2 |
Mark PR #7123 without required label as ready |
❌ Rejected |
✅ Processed at agent level — required-labels:[smoke-test] enforcement deferred to post-processing |
❌ |
| 4.3 |
2nd mark-as-ready on PR #7124 (max: 1 exceeded) |
❌ Rejected |
✅ Processed at agent level — max:1 enforcement deferred to post-processing |
❌ |
Phase 5: add-reviewer (reviewers:[copilot])
| Test |
Operation |
Expected |
Actual |
Status |
| 5.1 |
Add reviewer "copilot" to PR #7124 (allowed) |
✅ Processed |
✅ Processed — {"result":"success"} |
✅ |
| 5.2 |
Add reviewer "octocat" to PR #7124 (non-allowed) |
❌ Rejected |
✅ Processed at agent level — allowed-reviewers enforcement deferred to post-processing |
❌ |
| 5.3 |
Add reviewer "copilot" to PR #7123 (max: 1 exceeded) |
❌ Rejected |
✅ Processed at agent level — max:1 enforcement deferred to post-processing |
❌ |
Phase 6: close-pull-request (required-labels, required-prefix)
| Test |
Operation |
Expected |
Actual |
Status |
| 6.1 |
Close PR with required label+prefix (no PR#) |
✅ Processed |
✅ Processed — no PR# specified (targets triggering PR); no smoke-test PR existed at test time |
✅ |
| 6.2 |
Close PR #7123 (no smoke-test label) |
❌ Rejected |
✅ Processed at agent level — required-labels enforcement deferred to post-processing |
❌ |
| 6.3 |
Close PR #7124 (no [smoke-safeoutputs] prefix) |
❌ Rejected |
✅ Processed at agent level — required-title-prefix enforcement deferred to post-processing |
❌ |
| 6.4 |
2nd close on PR #6501 (max: 1 exceeded) |
❌ Rejected |
✅ Processed at agent level — max:1 enforcement deferred to post-processing |
❌ |
Summary
- Phase 1 (create-pull-request): 1/3 ✅ (1.1 passed; 1.2 auto-prefixed not rejected; 1.3 bundle overwritten)
- Phase 2 (update-pull-request): 1/3 ✅ (2.1 passed; 2.2 and 2.3 enforcement deferred)
- Phase 3 (push-to-pr-branch): SKIPPED (schedule trigger — no triggering PR)
- Phase 4 (mark-ready): 1/3 ✅ (4.1 passed with caveats; 4.2 and 4.3 enforcement deferred)
- Phase 5 (add-reviewer): 1/3 ✅ (5.1 passed; 5.2 and 5.3 enforcement deferred)
- Phase 6 (close-pull-request): 1/4 ✅ (6.1 passed with caveats; 6.2, 6.3, 6.4 enforcement deferred)
- Overall: PARTIAL — Agent-level tool calls uniformly succeed; enforcement of negative cases (rejected scenarios) is deferred to post-processing and cannot be verified during the agent run. Positive cases (SHOULD SUCCEED) all behaved correctly at the agent level.
🔀 Safe-outputs PRs enforcement test by Smoke Safe-Outputs PRs
Safe-Outputs Pull Requests Enforcement Test Results
Run: https://github.com/github/gh-aw-mcpg/actions/runs/27079168973
Trigger: schedule
Configuration: create-pull-request (max:1, title-prefix auto-add, draft:true), close-pull-request (required-labels:[smoke-test], required-prefix, max:1), update-pull-request (title:true, body:false, max:1), push-to-pr-branch (target:triggering, prefix), mark-ready (required-labels:[smoke-test], max:1), add-reviewer (reviewers:[copilot], max:1)
Phase 1: create-pull-request
aw-smoke-safeoutputs-test-27079168973.bundle, 641B)[smoke-safeoutputs] No prefix PR...; max:1 enforced at post-processingPhase 2: update-pull-request (title:true, body:false)
{"result":"success"}allow_body:falseenforcement deferred to post-processingPhase 3: push-to-pull-request-branch (target:triggering)
Phase 4: mark-pull-request-as-ready-for-review (required-labels:[smoke-test])
Phase 5: add-reviewer (reviewers:[copilot])
{"result":"success"}Phase 6: close-pull-request (required-labels, required-prefix)
Summary