What problem does it solve? After opening a pull request, developers must manually watch for CI failures, review comments, and merge conflicts. This Skill subscribes a session to a PR, receives pushed activity events from a background polling worker, and autonomously triages each event so the PR stays healthy until merge without constant user relaying. ## Core Features & Use Cases - Event-driven PR monitoring: Detects nine event types (CI failure/success, new comments, review approvals or change requests, merge conflicts, merged/closed) via wake prompts or advisory injection, each gated by configurable pr_monitor flags. - Structured triage taxonomy: Classifies every event as (a) clear low-risk fix routed through the swarm-pr-feedback discipline, (b) ambiguous or architecturally significant issue escalated to the user, or (c) informational noise acknowledged quietly. - Bounded retries and terminal cleanup: Applies a 3-strike rule per failing check before escalating with a diagnosis, and automatically ends subscriptions on pr.merged or pr.closed. - Use Case: After gh pr create opens a PR, the session auto-subscribes; when CI fails overnight, the worker wakes the session, the fix is verified, pushed, and reported — all without the user relaying the failure. ## Quick Start Subscribe this session to my open pull request and autonomously fix any CI failures or review comments until it merges.