# Using Pm Os

> Session guardrails for PM OS - dynamic execution with dependency awareness.

- Skill: `majiayu000/using-pm-os` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add majiayu000/using-pm-os`
- Raw SKILL.md: https://api.skillmd.com/api/skills/majiayu000/using-pm-os/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: majiayu000 (https://skillmd.com/u/majiayu000)
- Updated: 2026-09-09
- Page: https://skillmd.com/skills/majiayu000/using-pm-os

---


# Using PM OS

## Operating Principles

- PM OS is modular. Run only the skill you need for the question at hand.
- Dynamic flow is allowed, but dependency hygiene is required.

## Dependency Hygiene

Before generating any downstream output (charters or PRDs):
1. Read `alerts/stale-outputs.md`
2. If any outputs are stale, report them and ask whether to refresh
3. If a downstream output is newer than its sources, flag potential drift and ask to reconcile

## Evidence Discipline

Follow `CLAUDE.md` rules:
- Never invent metrics, quotes, or roadmap facts
- Tag claims as Evidence, Assumption, or Open Question
- Always include Sources Used and Claims Ledger in outputs