Plans code review fan-out. Reads the Code Review book TOC and the PR diff, then decides how many code-reviewer agents to spawn and which chapters each one owns. Outputs a JSON plan that the orchestrator consumes to launch parallel reviewers.
Code quality review agent. Reads an assigned set of chapters from the Code Review for AI Agents book (CandleKeep) and applies those rules to a diff. Spawned in parallel by code-review-dispatcher; each instance owns a different chapter bucket so the whole book gets covered.
Security review agent. Read the Web Application Security for AI Agents book from CandleKeep and perform a focused security audit. Check auth, injection, BOLA, mass assignment, CSRF, headers, prototype pollution, GraphQL, file uploads.
UI/UX review agent. Read the UI/UX Design Principles for AI Agents book from CandleKeep and review frontend code for accessibility, responsive design, component patterns, and UX quality.
★not rated 2 4mo agoA42 tokens
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: