Version 3.0 | AI-powered operating system for product managers
PM-OS transforms how product managers work by providing a unified, AI-native environment for context management, document generation, and workflow automation. It integrates with your existing tools (Jira, Slack, Google Workspace, GitHub) and provides a persistent "brain" that learns your projects, team, and preferences.
- Daily Context Sync - Automatically aggregate updates from Google Docs, Jira, GitHub, Slack, and Calendar into a single daily briefing
- Brain Knowledge Base - Persistent memory of entities (people, projects, teams, domains) with relationship tracking
- Session Persistence - Continue conversations across AI sessions with full context preservation
- Multi-Model Support - Works with Claude Code and Gemini CLI
- PRDs - Product Requirements Documents with First Principles Framework (FPF) reasoning
- ADRs - Architecture Decision Records
- RFCs - Request for Comments documents
- Meeting Notes - Structured meeting notes with action items
- PR/FAQs - Amazon-style working backwards documents
- Meeting Prep - Automatic pre-reads for upcoming calendar events
- Sprint Reports - Generate sprint summaries from Jira
- Context Updates - Daily synthesis of relevant information
- Slack Mention Capture - Track and triage mentions via bot
- First Principles Framework (FPF) - Structured reasoning with evidence tracking and decay monitoring
- Ralph - Multi-iteration feature development with acceptance criteria
- Confucius - Session note-taking for decisions, assumptions, and blockers
- Orthogonal Challenge - AI-powered devil's advocate for decisions
PM-OS uses a two-folder architecture separating logic from data:
pm-os/
├── common/ # LOGIC - shared code (this repo)
│ ├── .claude/commands/ # Claude Code slash commands (50+)
│ ├── .gemini/commands/ # Gemini CLI commands
│ ├── tools/ # Python tools and integrations
│ ├── frameworks/ # Document templates
│ ├── schemas/ # Entity schemas (YAML)
│ ├── rules/ # Agent behavior rules
│ ├── documentation/ # Full documentation
│ └── scripts/ # Boot and setup scripts
│
└── user/ # CONTENT - your data (separate repo)
├── brain/ # Knowledge base entities
│ ├── Entities/ # People, teams, domains
│ ├── Projects/ # Project context files
│ └── Experiments/ # Experiment tracking
├── context/ # Daily context files
├── sessions/ # Session persistence
├── planning/ # Meeting prep, career planning
├── config.yaml # Your configuration
└── .env # Your secrets
mkdir pm-os && cd pm-os
git clone https://github.com/feamando/pmos.git commonmkdir user
cp common/config.yaml.example user/config.yaml
cp common/.env.example user/.envEdit user/config.yaml with your details:
version: "3.0.0"
user:
name: "Your Name"
email: "your.email@company.com"
position: "Product Manager"
integrations:
jira:
enabled: true
url: "https://company.atlassian.net"
tracked_projects: ["PROJ1", "PROJ2"]
slack:
enabled: true
mention_bot_name: "your-slack-bot"
google:
enabled: trueEdit user/.env with your secrets:
# Jira
JIRA_API_TOKEN=your-api-token
# Slack
SLACK_BOT_TOKEN=xoxb-your-bot-token
# Google (OAuth paths)
GOOGLE_CREDENTIALS_PATH=.secrets/credentials.json
GOOGLE_TOKEN_PATH=.secrets/token.jsonpip install -r common/requirements.txtcd user
source ../common/scripts/boot.shIn your AI CLI (Claude Code or Gemini), run:
/boot
PM-OS includes 50+ slash commands organized by category:
| Command | Description |
|---|---|
/boot |
Initialize PM-OS and load context |
/update-context |
Sync daily context from all integrations |
/session-save |
Save current session for later |
/session-load |
Resume a previous session |
| Command | Description |
|---|---|
/prd |
Generate a PRD with FPF reasoning |
/adr |
Create an Architecture Decision Record |
/rfc |
Create a Request for Comments |
/meeting-notes |
Structure meeting notes |
/prfaq |
Generate a PR/FAQ document |
| Command | Description |
|---|---|
/jira-sync |
Sync projects from Jira |
/github-sync |
Sync repositories from GitHub |
/confluence-sync |
Sync spaces from Confluence |
/slack-sync |
Sync channels from Slack |
| Command | Description |
|---|---|
/q0-init |
Initialize reasoning context |
/q1-hypothesize |
Generate competing hypotheses |
/q2-verify |
Verify hypotheses logically |
/q3-validate |
Validate with evidence |
/q4-audit |
Calculate trust scores |
/q5-decide |
Create decision record |
/q-status |
Show FPF state |
| Command | Description |
|---|---|
/meeting-prep |
Prepare for upcoming meetings |
/sprint-report |
Generate sprint summary |
/ralph-init |
Start multi-iteration feature |
/ralph-status |
Check feature progress |
- Gmail - Email context extraction
- Google Docs - Document sync and analysis
- Google Calendar - Meeting prep automation
- Google Drive - File discovery
- Jira - Issue tracking, sprint data
- Confluence - Documentation sync
- Slack - Channel monitoring, mention capture, context posting
- GitHub - Repository activity, PR tracking
- Statsig - Experiment tracking (optional)
The Brain is PM-OS's persistent memory, storing structured entities:
- Person - Team members, stakeholders, contacts
- Team - Squads, tribes, organizational units
- Project - Initiatives, features, products
- Domain - Technical domains, system areas
- Experiment - A/B tests, pilots, trials
All entities follow a consistent YAML frontmatter structure:
---
id: entity-id
type: project
title: "Project Name"
status: active
owner: person-id
created: 2026-01-01
last_updated: 2026-01-15
related:
- "[[Entities/Related_Person.md]]"
- "[[Projects/Related_Project.md]]"
---
# Entity content in markdown...The Brain loader identifies frequently-mentioned entities in your daily context and surfaces them as "hot topics" for quick reference.
Full configuration options:
version: "3.0.0"
user:
name: "Your Name"
email: "email@company.com"
position: "Product Manager"
tribe: null # Optional: your tribe/team
team: null # Optional: your squad
integrations:
jira:
enabled: false
url: "https://company.atlassian.net"
username: ""
tracked_projects: []
default_project: ""
github:
enabled: false
org: ""
tracked_repos: []
slack:
enabled: false
mention_bot_name: ""
google:
enabled: false
confluence:
enabled: false
url: ""
spaces: []
pm_os:
fpf_enabled: true
confucius_enabled: true
ralph_enabled: true
auto_update: false
brain:
entity_types: [person, team, project, domain, experiment]
hot_topics_limit: 10
validate_on_load: true
context:
retention_days: 30
include:
jira: true
github: true
slack: true
calendar: true
sessions:
auto_save_interval: 30
max_sessions: 50See .env.example for all available environment variables.
If upgrading from PM-OS v2.4:
/update-3.0
This will:
- Create a backup snapshot
- Migrate Brain, sessions, and context
- Generate config.yaml from existing setup
- Verify the migration
See Migration Guide for details.
PM-OS includes a standalone developer toolkit for "vibe coding" that can be extracted to any project:
- Session management
- Confucius note-taking
- Ralph feature tracking
- FPF reasoning
See Developer Tools for standalone extraction.
- Python 3.9+
- Git
- Claude Code or Gemini CLI
- (Optional) Jira, GitHub, Slack accounts for integrations
Full documentation is available in the documentation/ directory:
- Overview
- Architecture
- Installation
- Workflows
- Brain System
- Command Reference
- Tool Reference
- Troubleshooting
Contributions are welcome! Please:
- Fork the repository
- Create a feature branch
- Submit a pull request
See CONTRIBUTING.md for guidelines.
MIT License - See LICENSE
PM-OS v3.0 - AI-powered operating system for product managers