rai skill set list

Systematic root cause analysis using lean methods (5 Whys, Ishikawa, Gemba). Use when encountering unexpected behavior, errors, or defects to find and fix the true root cause rather than symptoms.

Run the full codebase discovery pipeline: detect languages, extract symbols, describe components, generate architecture docs, and build the knowledge graph.

Compare knowledge graph against module architecture docs and update drifted fields. Deterministic frontmatter comparison using existing rai graph commands, with inference for narrative sections. HITL before any writes.

Design an epic from strategic objective to feature breakdown. Use when starting a new body of work spanning multiple features (3-10), requiring architectural decisions, or when establishing technical direction for significant capability delivery.

Guided MCP server registration. Collects intent conversationally, resolves package details, and delegates to `rai mcp install` or `rai mcp scaffold`. Human never constructs CLI commands.

Safe MCP server removal with adapter dependency checking. Shows registered servers, warns about references, deletes config.

Guided problem definition at portfolio level. Takes a vague business idea and shapes it into a well-formed problem statement before it enters the epic pipeline. Produces a Problem Brief that feeds /rai-epic-design.

Guide greenfield project setup through conversation. Fills governance templates with project-specific content and builds the knowledge graph. Use after rai init on a new project.

Guide brownfield project onboarding through discovery and conversation. Analyzes existing codebase, detects conventions, fills governance templates with discovered and conversational content, and builds the knowledge graph. Use after rai init --detect on an existing project.

Conduct epistemologically rigorous research to inform decisions. Use before ADRs, when evaluating competing approaches, entering unfamiliar domains, or resolving parking lot items. Produces evidence catalogs with triangulated claims and actionable recommendations.

Close a working session by reflecting on outcomes and feeding structured data to CLI. CLI does all writes atomically; skill does inference reflection.

Begin a session by loading context bundle, interpreting it, and proposing work. CLI does all data plumbing; skill does inference interpretation.

Complete a story with retrospective verification, local merge to dev, and tracking update. MRs are created at epic level, not per story. Use after review to formally close the story lifecycle.

Execute the implementation plan task by task, verifying each step, and producing quality code that passes validation gates. Use after planning is complete.

Decompose user stories into atomic executable tasks, identify dependencies, and create a deterministic implementation plan. Use after /rai-story-design has grounded the story's integration decisions.

Reflect on completed stories to extract learnings, identify process improvements, and update the framework with insights gained. Use after implementation is complete to close the development cycle.

Initialize a story with verified context, branch, and scope commit. Use at the beginning of story work to ensure proper setup and traceability from the start.