Repository navigation
feat: add composite task routing eval - #61
Conversation
|
Caution Review failedThe pull request is closed. ℹ️ Recent review info⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (3)
📝 WalkthroughWalkthroughAdaptive routing is updated to version 1.9.1 with composite task decomposition, dependency-ordered route merging, and selective skill loading. A new evaluation scenario verifies debugging, regression, and boundary-review routing behavior. ChangesAdaptive routing composition
Estimated code review effort: 3 (Moderate) | ~20 minutes Possibly related issues
Possibly related PRs
Sequence Diagram(s)sequenceDiagram
participant Agent
participant AdaptiveRouting
participant DiscoveryMetadata
participant SpecialistSkills
Agent->>AdaptiveRouting: Decompose CI-failure request
AdaptiveRouting->>DiscoveryMetadata: Match debugging, regression, and boundary work
DiscoveryMetadata-->>AdaptiveRouting: Return candidate paths and skills
AdaptiveRouting->>SpecialistSkills: Load selected specialist skills in order
AdaptiveRouting-->>Agent: Report merged route and selected files
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
There was a problem hiding this comment.
🧹 Nitpick comments (1)
directives/adaptive-routing.md (1)
108-110: 🎯 Functional Correctness | 🔵 Trivial | ⚡ Quick winClarify regression-test ordering.
“Fix before regression test” can encourage adding the regression test only after the defect is fixed, so it never demonstrates the original failure. Prefer: reproduce → add/update regression test → fix → rerun the test → boundary review. Align the scenario’s route expectations with this distinction.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@directives/adaptive-routing.md` around lines 108 - 110, Clarify the route-ordering guidance in the “Compose the final route” step: require reproduce → add or update the regression test → fix → rerun the test → boundary review, replacing the ambiguous “fix before regression test” ordering and aligning scenario route expectations accordingly.
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Nitpick comments:
In `@directives/adaptive-routing.md`:
- Around line 108-110: Clarify the route-ordering guidance in the “Compose the
final route” step: require reproduce → add or update the regression test → fix →
rerun the test → boundary review, replacing the ambiguous “fix before regression
test” ordering and aligning scenario route expectations accordingly.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: 3b6e458d-9ef9-40e4-aa95-6195a38159bb
📒 Files selected for processing (3)
directives/adaptive-routing.mdevals/scenarios/compositional-skill-routing.mdmanifest.json
Summary
directives/adaptive-routing.md.manifest.jsonafter bumpingadaptive-routingto 1.9.0.Verification
npm install(dependencies were missing locally)npm run manifestgit diff --check && git diff --cached --checknpm run checkgit diff --check HEAD~1..HEADParallel PR notes
This PR intentionally stays scoped to routing prose and an eval scenario. It does not add manifest routing metadata or context-audit CLI behavior, which are expected to be handled in separate parallel PRs.
Summary by CodeRabbit
New Features
Documentation