AI-Assisted Development Workflow for planning, tasking, implementing, auditing project implementations with continuously self-improving retrospectives.
Agent Claude Code
Pair-programming engineer for architecture, test-driven development, refactoring, and code quality. Use when building, designing, or improving code, including driving red/green/refactor cycles via /tdd-pair.
Agent Claude Code
Research specialist for gathering and synthesizing technical information from the web and documentation. Use when a question needs sourced research.
Agent Claude Code
UX design specialist for accessible interaction patterns and component specs. Use when a problem needs expert UX design thinking.
Agent Claude Code
Discovery and planning analyst for code archaeology, solution design, and risk analysis. Use when starting a discovery task or planning an implementation.
Agent Claude Code
Principal security engineer for threat modeling, security architecture review, and risk assessment. Use for security reviews.
Agent Claude Code
Task generation and sprint planning specialist. Use to break work into estimable tasks and assign story points under the Task Workflow System.
Agent Claude Code
Principal-level technical reviewer. Use to review code, specs, or implementation phases for correctness, security, performance, and maintainability, ending with an explicit verdict.
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: