Problem
When an agent reviews its own PR (authored in the same session), the review reads in third-person congratulatory style ("Great job! Thanks for putting this together! LGTM!"). This is semantically incorrect — the agent is praising its own work.
Solution
Add a self-review detection heuristic to the pr-review skill that switches the review tone based on authorship:
Detection Logic
- Extract the ticket number from the PR body (
Resolves #N)
- Query
query_raw_memories scoped to the current session ID for references to that ticket
- Match → Agent authored this PR in same session → self-review mode (first-person, clinical)
- No match → Peer-review mode (third-person, constructive)
Author Matrix
| Author |
Reviewer |
Session Match |
Mode |
| Agent (session A) |
Same agent (session A) |
✅ |
Self-review |
| Agent (session A) |
Different agent (session B) |
❌ |
Peer-review |
| Human |
Agent |
❌ |
Peer-review |
| External contributor |
Agent |
❌ |
Peer-review |
File Changes
pr-review-guide.md — Add self-review bullet to Section 1 (Core Philosophy), add detection heuristic to Section 2 (Agent Mandates)pr-review-template.md — Add alternative self-review opening
Origin Session ID: 144326a4-c1e8-4eee-80e4-b984f3c79ddc
Problem
When an agent reviews its own PR (authored in the same session), the review reads in third-person congratulatory style ("Great job! Thanks for putting this together! LGTM!"). This is semantically incorrect — the agent is praising its own work.
Solution
Add a self-review detection heuristic to the
pr-reviewskill that switches the review tone based on authorship:Detection Logic
Resolves #N)query_raw_memoriesscoped to the current session ID for references to that ticketAuthor Matrix
File Changes
pr-review-guide.md— Add self-review bullet to Section 1 (Core Philosophy), add detection heuristic to Section 2 (Agent Mandates)pr-review-template.md— Add alternative self-review openingOrigin Session ID: 144326a4-c1e8-4eee-80e4-b984f3c79ddc