pmm-onboard

What problem does it solve?

PMM needs a reliable, privacy-safe identity layer for every new agent, but onboarding through other AIs or outdated profiles leaves gaps, so this skill captures operative identity and routes sensitive data before work begins.

Core Features & Use Cases

• Extraction mode: Provide the previous AI's summary so the skill maps each section into user.md and secrets.md, asks targeted follow-ups for gaps, and commits the identity file.
• Interview mode: Run five to seven focused questions when no prior AI exists, translating the answers into operative communication, cognitive, rhythm, and principle sections along with PII routing.
• Refresh mode & gating: Diff current user.md, offer re-extract or re-interview paths, respect org-wide onboard_agent restrictions, and teach the maintain cycle to append calibration or anti-pattern notes without changing base sections.

Quick Start

Provide the extraction output or interview answers while seeding the user identity layer so the assistant writes user.md and secrets.md through PMM's onboarding flow.