Adversarial swarm PR code review. Fetches diff, sweeps for relevant concerns, fans out per-file validators, consolidates findings, and posts inline GitHub comments.
Run adversarial swarm review then auto-fix all findings in parallel. Use when the user says "review and fix", "swarm fix", or wants review and fix in one shot.
Fetch every PR review thread (Copilot and human), triage each one, fix them in parallel, verify with happy/sad-path coverage, refresh the PR title and description, then reply to and resolve every thread. Use when the user says "address the review threads",…
Execute a workflow definition. Parses YAML, compiles DAG, walks parallel groups, spawns agents per stage, handles human gates and validation.
Split an oversized Python file into smaller focused modules, updating all call sites directly (no re-export facades). Use when a file is too large to review or maintain comfortably.
Shared project rules for all dancing-bear agents. Covers CLI conventions, credential protection, output helpers, coding standards, and the plan→dry-run→apply pattern.
Diagnose and fix GitHub Actions CI failures. Given a URL, PR number, or "CI is failing", parses logs, identifies root cause, fixes locally, and verifies. Use when CI is red and you need to get it green.
Analyze test coverage gaps on changed or specified files, spawn tester agents to write tests, and verify improvement. Use when you need to get coverage above 80% on new code or expand coverage on existing modules.