pm-augmente
A starter pack of 12 skills and 6 subagents for Product Managers.
Open source Repository Open in the app JSON README (API)
About
A starter pack of 12 skills and 6 subagents for Product Managers.
Details
- Kind
- Plugins
- Topic
- No topic detected
- Publisher
- ludovic33fr
- Origin
- marketplace
- Category
- ferramentas
- Stars
- 1
- Last push
- 2026-04-30T22:20:12Z
- Repository state
- ativo
- Language
- Python
- License
- MIT
- Added
- 2026-08-30 01:48:58
- Updated
- 2026-08-30 01:48:58
- Origin id
ludovic33fr/product-ai-toolbox/pm-augmente
README
# pm-augmente — Augmented Product Manager Toolbox > Starter pack of **12 skills + 6 agents** for [Claude Code](https://claude.com/claude-code), companion to the book *Le Product Manager Augmenté* by Ludovic Lefebvre. [](LICENSE) [](.claude-plugin/plugin.json) [](README.fr.md) > **Heads-up.** Skill and agent bodies, prompts and produced outputs are in **French**. The repository envelope (this README, contribution rules, frontmatter keys) is in English to maximize international discoverability. French speakers should start with [README.fr.md](README.fr.md). --- ## Install ### Option 1 — As a Claude Code plugin (recommended) ```bash /plugin marketplace add Ludovic33Fr/product-ai-toolbox /plugin install pm-augmente@product-ai-toolbox ``` ### Option 2 — Manual copy Copy the folders directly into your Claude Code config: ```bash cp -r skills/* ~/.claude/skills/ cp -r agents/* ~/.claude/agents/ ``` See [docs/installation.md](docs/installation.md) (FR) for variants (Cursor, cron, scheduled invocations). --- ## What's inside ### Skills (12) | Name | One-line gloss (EN) | |------|---------------------| | `weak-signals-weekly` | Weekly weak-signals synthesis from verbatims, support, mentions, logs | | `quality-review-tickets` | Pre-refinement quality review of backlog tickets (INVEST + edge cases) | | `discovery-cadrage` | Frame a user discovery with hypotheses, target profile, interview guide | | `committee-prep` | Prepare an arbitration committee with anticipated objections and decision criteria | | `spec-from-brief` | Turn a fuzzy brief into a testable, exhaustive specification | | `roadmap-variants` | Simulate three roadmap variants (cautious / balanced / aggressive) | | `multi-stakeholder-comm` | Tailor one topic into stakeholder-specific one-pagers | | `onboarding-pm` | Build the onboarding path for a new PM joining the team | | `devils-advocate` | Have the AI play professional skeptic to demolish a proposal | | `feature-ethics-audit` | Audit a planned feature for ethical risks (manipulation, bias, data) | | `quarterly-review-pm` | Quarterly PM review based on factual deliveries and AI maturity grid | | `four-indicators-measure` | Compute the four augmented-PM indicators | ### Agents (6) | Name | One-line gloss (EN) | |------|---------------------| | `competitive-watcher` | Continuous competitive watch + weekly 5-minute briefing note | | `backlog-triage-agent` | Pre-classify incoming tickets to prepare human decision | | `roadmap-pilot-agent` | Compare roadmap progress vs plan, weekly steering note | | `weak-signals-watcher` | Continuously detect weak user signals, feed the discovery journal | | `decision-journal-keeper` | Keep the PM decision journal up to date without manual capture | | `stakeholder-digest-agent` | Weekly per-stakeholder one-pager drafts, ready for human review | Detailed mapping with the book chapters: [docs/livre.md](docs/livre.md) (FR). --- ## The book This repository is the public companion of: > **Le Product Manager Augmenté — Décider mieux, plus vite, avec l'IA** > Ludovic Lefebvre, 2026. The book introduces the augmented PM model, refounds operational gestures (discovery, decision, specification, alignment), and provides a 90-day transformation plan. This repo turns Annex 3 into executable artefacts. Book landing page (link added at publication). --- ## Contributing We welcome new skills, agents, and improvements. See [CONTRIBUTING.md](CONTRIBUTING.md) (EN) or [CONTRIBUTING.fr.md](CONTRIBUTING.fr.md) (FR — editorial criteria from the book). Three acceptance criteria, lifted from the book: 1. **Documentation** — required fields complete (4 for skills, 5 for agents). 2. **Versioning** — author, date, motivation in the frontmatter. 3. **Guardrails** — explicit limits, when not to use. A monthly community review merges contributions, retires obsolete artefacts, and ships a new release. --- ## License [MIT](LICENSE) © 2026 Ludovic Lefebvre.