onboard

Role-based onboarding documentation is generated by scanning project directories and conventions to produce a focused Markdown summary. It reads project standards and optional role definitions to tailor the output for specific contributor areas like programming or QA.

Onboarding a new agent involves summarizing project state, architecture, and current priorities into a structured document. The process scans relevant directories based on the agent's specified role and incorporates recent change context to provide actionable next steps.

Yes, separate onboarding documents can be created for QA, design, programming, narrative, and production contexts. The generation process locates and summarizes key directories specific to each role, such as scanning tests for QA or src for programmers.

Prompt-driven contributor enablement works by reading project overview files like CLAUDE.md and role-specific definitions to generate structured Markdown. It produces first tasks, pitfalls, and guiding questions tailored to the specified role.

You need a CLAUDE.md file to provide project overview standards for the onboarding generation. Optionally, role definitions located under the .claude/agents/ directory can be read to further tailor the generated onboarding output.

The generated role-based onboarding document can optionally be written directly to a production onboarding path. This allows the structured Markdown output to be reused for future contributors entering the project.