pm-claude-skills 80.1.0 → 80.2.0

package/README.md

@@ -10,22 +10,25 @@
   </a>
 </p>
 
+<p align="center">
+  <a href="#-quick-start"><strong>Quick start</strong></a> &middot;
+  <a href="https://mohitagw15856.github.io/pm-claude-skills/"><strong>Playground</strong></a> &middot;
+  <a href="SKILLS.md"><strong>All skills</strong></a> &middot;
+  <a href="PACKS.md"><strong>Packs</strong></a> &middot;
+  <a href="docs/installation.md"><strong>Install guide</strong></a> &middot;
+  <a href="CHANGELOG.md"><strong>Changelog</strong></a> &middot;
+  <a href="ROADMAP.md"><strong>Roadmap</strong></a>
+</p>
+
 <p align="center">
   <a href="https://github.com/mohitagw15856/pm-claude-skills/stargazers"><img src="https://img.shields.io/github/stars/mohitagw15856/pm-claude-skills?style=social" alt="Stars"></a>
   <a href="SKILLS.md"><img src="https://img.shields.io/badge/dynamic/json?url=https%3A%2F%2Fmohitagw15856.github.io%2Fpm-claude-skills%2Fskills.json&query=%24.count&label=skills&color=blue" alt="Skills"></a>
   <a href="https://github.com/mohitagw15856/pm-claude-skills/releases"><img src="https://img.shields.io/github/v/release/mohitagw15856/pm-claude-skills?label=version&color=brightgreen" alt="Version"></a>
   <a href="https://www.npmjs.com/package/pm-claude-skills"><img src="https://img.shields.io/npm/v/pm-claude-skills?logo=npm&color=cb3837" alt="npm"></a>
-  <a href="https://pypi.org/project/pm-skills/"><img src="https://img.shields.io/pypi/v/pm-skills?logo=pypi&logoColor=white&color=3775A9&label=pip" alt="PyPI"></a>
   <a href="#-quick-start"><img src="https://img.shields.io/badge/Anthropic%20Plugin%20Directory-Published-D97757?logo=anthropic&logoColor=white" alt="In the official Anthropic plugin directory"></a>
-  <br>
   <a href=".github/workflows/skillcheck.yml"><img src="https://img.shields.io/github/actions/workflow/status/mohitagw15856/pm-claude-skills/skillcheck.yml?branch=main&label=SkillCheck" alt="SkillCheck"></a>
-  <a href="conformance/REGISTRY.md"><img src="https://img.shields.io/endpoint?url=https%3A%2F%2Fraw.githubusercontent.com%2Fmohitagw15856%2Fpm-claude-skills%2Fmain%2Fconformance%2Fbadge.json" alt="SkillSpec"></a>
   <a href=".github/workflows/skill-audit.yml"><img src="https://img.shields.io/github/actions/workflow/status/mohitagw15856/pm-claude-skills/skill-audit.yml?branch=main&label=security%20audit" alt="Security Audit"></a>
-  <a href="https://pm-skills-mcp.pm-claude-skills.workers.dev/today.json"><img src="https://img.shields.io/endpoint?url=https%3A%2F%2Fpm-skills-mcp.pm-claude-skills.workers.dev%2Ftoday%2Fbadge" alt="Skill of the day"></a>
-  <a href="https://mohitagw15856.github.io/pm-claude-skills/"><img src="https://img.shields.io/endpoint?url=https%3A%2F%2Fpm-skills-mcp.pm-claude-skills.workers.dev%2Ftry%2Fstats" alt="Free runs served"></a>
-  <a href="https://github.com/BehiSecc/awesome-claude-skills"><img src="https://img.shields.io/badge/Awesome%20Claude%20Skills-listed-fc60a8?logo=awesomelists&logoColor=white" alt="Listed in Awesome Claude Skills"></a>
   <a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-lightgrey" alt="License"></a>
-  <a href="https://github.com/sponsors/mohitagw15856"><img src="https://img.shields.io/badge/sponsor-❤️-ff69b4" alt="Sponsor"></a>
 </p>
 
 > **Your landlord kept your deposit. Your mom got a medical bill that makes no sense. You got laid off on a Tuesday. Someone you love died, and no one handed you the checklist.**
@@ -38,7 +41,7 @@ PM Skills is an open-source library of 1176 Agent Skills — plain-markdown SKIL
 
 <table align="center"><tr>
 <td align="center"><b>1176</b><br><sub>skills</sub></td>
-<td align="center"><b>132</b><br><sub>bundles</sub></td>
+<td align="center"><b>134</b><br><sub>bundles</sub></td>
 <td align="center"><b>35</b><br><sub>professions</sub></td>
 <td align="center"><b>12</b><br><sub>platforms</sub></td>
 <td align="center"><b>4.8 / 5</b><br><sub><a href="https://mohitagw15856.github.io/pm-claude-skills/leaderboard.html">eval-scored</a></sub></td>
@@ -46,6 +49,14 @@ PM Skills is an open-source library of 1176 Agent Skills — plain-markdown SKIL
 <td align="center"><b>MIT</b><br><sub>forever</sub></td>
 </tr></table>
 
+## 🧭 How it works
+
+| | Step | Example |
+|---|---|---|
+| **01** | Say what you need, in your own words | *"My landlord is keeping my deposit."* |
+| **02** | Your assistant loads the one skill that matches | [`security-deposit-recovery`](skills/security-deposit-recovery/SKILL.md), a single markdown file, read only when it is relevant |
+| **03** | You get the finished document, not advice about it | The challenge to each deduction, the demand letter, and the point at which small claims is worth it |
+
 ## 🚪 Pick a door — 30 seconds each
 
 | | You… | Do this |
@@ -58,25 +69,26 @@ PM Skills is an open-source library of 1176 Agent Skills — plain-markdown SKIL
 
 **Nothing here can scare your setup.** A skill is a markdown file your AI reads. No runtime, no telemetry, no accounts. Installing copies text files; uninstalling is deleting them. Skeptical? Good instinct — [read one first](skills/lease-decoder/SKILL.md), it's written for humans too.
 
-<p align="center">
-  <a href="docs/installation.md"><img src="https://img.shields.io/badge/Claude_Code-native-D97757?style=for-the-badge&logo=anthropic&logoColor=white" alt="Claude Code — native"></a>
-  <a href="exports/chatgpt/"><img src="https://img.shields.io/badge/ChatGPT-exports-74aa9c?style=for-the-badge&logo=openai&logoColor=white" alt="ChatGPT exports"></a>
-  <a href="exports/gemini/"><img src="https://img.shields.io/badge/Gemini-exports-4285F4?style=for-the-badge&logo=googlegemini&logoColor=white" alt="Gemini exports"></a>
-  <a href="docs/installation.md"><img src="https://img.shields.io/badge/Cursor_·_Codex_·_Windsurf-one_command-1a1a2e?style=for-the-badge" alt="Cursor, Codex, Windsurf — one command"></a>
-  <a href="mcp-remote/"><img src="https://img.shields.io/badge/MCP-any_client-8a5cf5?style=for-the-badge" alt="MCP — any client"></a>
-  <br>
-  <a href="integrations/telegram/"><img src="https://img.shields.io/badge/Telegram-bot-26A5E4?style=for-the-badge&logo=telegram&logoColor=white" alt="Telegram bot"></a>
-  <a href="integrations/slack-app/"><img src="https://img.shields.io/badge/Slack-%2Fskill-4A154B?style=for-the-badge&logo=slack&logoColor=white" alt="Slack app"></a>
-  <a href="integrations/raycast/"><img src="https://img.shields.io/badge/Raycast-launcher-FF6363?style=for-the-badge&logo=raycast&logoColor=white" alt="Raycast launcher"></a>
-  <a href="integrations/obsidian-plugin/"><img src="https://img.shields.io/badge/Obsidian-plugin-7C3AED?style=for-the-badge&logo=obsidian&logoColor=white" alt="Obsidian plugin"></a>
-  <a href="connectors/"><img src="https://img.shields.io/badge/n8n-connector-EA4B71?style=for-the-badge&logo=n8n&logoColor=white" alt="n8n connector"></a>
-  <a href="integrations/jev/"><img src="https://img.shields.io/badge/Jev-decision_layer-0f9d58?style=for-the-badge" alt="Jev decision layer"></a>
-  <br>
-  <a href="https://pypi.org/project/pm-skills/"><img src="https://img.shields.io/badge/Python-pip_install_pm--skills-3776AB?style=for-the-badge&logo=python&logoColor=white" alt="Python — pip install pm-skills"></a>
-  <a href="dataset/"><img src="https://img.shields.io/badge/🤗_Hugging_Face-dataset-FFD21E?style=for-the-badge" alt="Hugging Face dataset"></a>
-  <a href="Dockerfile"><img src="https://img.shields.io/badge/Docker-ghcr_image-2496ED?style=for-the-badge&logo=docker&logoColor=white" alt="Docker image on ghcr"></a>
-  <a href="action/"><img src="https://img.shields.io/badge/GitHub_Actions-CI_skills-2088FF?style=for-the-badge&logo=githubactions&logoColor=white" alt="GitHub Actions"></a>
-</p>
+<div align="center">
+
+| Works with | |
+|---|---|
+| **Assistants** | [Claude Code](docs/installation.md) · [ChatGPT](exports/chatgpt/) · [Gemini](exports/gemini/) · [Cursor, Codex, Windsurf](docs/installation.md) · [any MCP client](mcp-remote/) |
+| **Where you already work** | [Telegram](integrations/telegram/) · [Slack](integrations/slack-app/) · [Raycast](integrations/raycast/) · [Obsidian](integrations/obsidian-plugin/) · [n8n](connectors/) |
+| **For builders** | [Python](https://pypi.org/project/pm-skills/) · [Hugging Face dataset](dataset/) · [Docker](Dockerfile) · [GitHub Actions](action/) · [decision layer](integrations/jev/) |
+
+<em>If it can read a markdown file, it can use a skill.</em>
+
+</div>
+
+## ✅ PM Skills is right for you if
+
+- ✅ You keep retyping the same long prompt and get a different answer each time
+- ✅ You want **the finished document** (a PRD, a demand letter, a postmortem), not an essay on how to write one
+- ✅ You use **more than one assistant** and want the same skills in all of them
+- ✅ You want to **read exactly what your assistant was told** before you trust the result
+- ✅ You are dealing with something outside your field: a lease, a medical bill, a layoff
+- ✅ You want nothing to host, no account, and no telemetry unless you switch it on
 
 ## 🥊 Without a skill vs. with one
 
@@ -107,82 +119,16 @@ PM Skills is an open-source library of 1176 Agent Skills — plain-markdown SKIL
 </tr>
 </table>
 
-## 🎨 New: design taste
+## 🆕 What's new
 
-Ask an assistant for a landing page and you get the same one everybody gets. The **[pm-design-taste](plugins/pm-design-taste/)** bundle fixes the order of decisions instead of adding more rules: `ui-design-pipeline` runs five stages in sequence and hands each one to a specialist skill if you have it installed.
-
-| Stage | What happens | Hands off to | Pointer skill |
+| Release | What it adds | Try saying | Read more |
 |---|---|---|---|
-| 1. Direction | Commit to an aesthetic stance from audience and purpose, before any code | frontend-design (Anthropic) | [`frontend-design-pointer`](skills/frontend-design-pointer/SKILL.md) |
-| 2. System | Lock the direction into a persisted design system file so later pages do not drift | ui-ux-pro-max (nextlevelbuilder) | [`ui-ux-pro-max-pointer`](skills/ui-ux-pro-max-pointer/SKILL.md) |
-| 3. Motion | Enter animations ease out, most motion under 300ms, animate only where it helps | emil-design-eng (Emil Kowalski) | [`emil-design-eng-pointer`](skills/emil-design-eng-pointer/SKILL.md) |
-| 4. Critique | Hierarchy, spacing, typography, and "too safe vs too loud" | impeccable (Paul Bakaus) | [`impeccable-pointer`](skills/impeccable-pointer/SKILL.md) |
-| 5. Verification | Real browser, mobile and desktop widths, screenshot each step, fix and re-check | playwright-cli (Microsoft) | [`playwright-cli-pointer`](skills/playwright-cli-pointer/SKILL.md) |
-
-The orchestrator is [`ui-design-pipeline`](skills/ui-design-pipeline/SKILL.md). If a specialist skill is missing it uses its own built-in guidance and tells you the install command. It never loads all the design skills at once, because they give conflicting instructions.
-
-```
-/plugin install pm-design-taste@pm-claude-skills
-```
-
-Try it:
-
-- "Run the design pipeline on this landing page."
-- "This dashboard looks AI-generated. Fix it properly."
-- "Which design skills do I have installed, and what is missing for the pipeline?"
-- "Check the signup page in a real browser at mobile and desktop widths."
+| **[v80.2.0](https://github.com/mohitagw15856/pm-claude-skills/releases/latest)** trust in a release | The MCP install command fixed, a check that every release installs and runs, announcements in `doctor` and the playground, and frozen versions of the flagship skills that a team can pin | *"Run the doctor."* | [CHANGELOG.md](CHANGELOG.md) · [skill releases](docs/SKILL-RELEASES.md) |
+| **[v80.1.0](https://github.com/mohitagw15856/pm-claude-skills/releases/tag/v80.1.0)** design taste | A five-stage UI pipeline (direction, system, motion, critique, verification) that stops generic AI-looking interfaces, with pointers to five specialist design skills | *"Run the design pipeline on this landing page."* | [pm-design-taste](plugins/pm-design-taste/) · [credits](plugins/pm-design-taste/THIRD_PARTY.md) |
+| **v80.0.0** the promote loop | Scans your own transcripts for what you keep asking, drafts it as a skill, tests the triggers, publishes it | *"What do I keep asking you for?"* | [PROMOTE-LOOP.md](docs/PROMOTE-LOOP.md) |
+| **v79.0.0** the decision layer | Typed questions with a probability per option: which skill, is this input safe, ship or slip | *"Should we ship Friday?"* | [JEV-DECISION-LAYER.md](docs/JEV-DECISION-LAYER.md) |
 
-The five specialist skills are third-party projects. Nothing of theirs is copied here: the pointers give the install command and credit the author. Authors, licences and links are in [`THIRD_PARTY.md`](plugins/pm-design-taste/THIRD_PARTY.md). More prompts in [`examples/pm-design-taste-example.md`](examples/pm-design-taste-example.md).
-
-## ♻️ New: promote what you keep asking for
-
-You have five prompts you type every week with small changes. Each one is a skill that hasn't been written down yet. The **[pm-skill-promoter](plugins/pm-skill-promoter/)** bundle closes the loop: scan your own transcripts, pick a pattern, draft the skill, test its triggers, publish it. One command: `/promote ~/.claude/projects/my-project`.
-
-Here is the loop on the fixture in [`examples/promoter/`](examples/promoter/), a made-up three-week transcript with a few secrets planted in it:
-
-```
-$ python3 skills/promoter-scan/scripts/promoter_scan.py examples/promoter
-
-| # | Pattern                            | Times | Score |
-|---|------------------------------------|------:|------:|
-| 1 | write release notes log            |    14 |    82 |   ← you asked for this FOURTEEN times
-| 2 | draft update notes stakeholder     |     6 |    71 |
-| 3 | summarise thread decisions owners  |     5 |    60 |
-
-  "my api key is [api-key] please use it"      ← the planted secrets come out redacted
-  "email the report to [email] and cc [phone]"
-```
-
-Pattern 1 becomes [`release-notes-from-git-log`](examples/promoter/drafted-skill/SKILL.md): a trigger built from the phrases you actually typed, three worked examples with your product name removed, and the corrections you kept making ("shorter", "group by feature", "user-facing tone") turned into quality checks. Then the test:
-
-```
-$ python3 skills/promoter-test/scripts/promoter_test.py drafted-skill/SKILL.md drafted-skill/evals.json
-precision 1.0  recall 1.0        ← fires on "changelog entry from these commits", stays quiet on "write a haiku about deploys"
-```
-
-Then [`promoter-publish`](skills/promoter-publish/SKILL.md) prints the entire release package and asks once before pushing. The whole story, with every file: **[docs/PROMOTE-LOOP.md](docs/PROMOTE-LOOP.md)**. Nothing leaves your machine during the scan; the redaction runs before anything is written.
-
-## 🎯 New: the decision layer
-
-<p align="center">
-  <a href="docs/JEV-DECISION-LAYER.md"><picture><source media="(prefers-color-scheme: dark)" srcset="web/docs-assets/decision-layer.svg"><source media="(prefers-color-scheme: light)" srcset="web/docs-assets/decision-layer-light.svg"><img src="web/docs-assets/decision-layer.svg" width="100%" alt="The decision layer: a prompt goes in, a calibrated decision model picks the pack then the skill with probabilities, the skill runs, the artifact comes out — and the same typed questions guard inputs, gate journeys and judge evals" /></picture></a>
-</p>
-
-Some questions deserve a **number, not a paragraph**: *which of 1176 skills?* · *is this input safe?* · *escalate or hold?* · *does this output clear the bar?* The library now answers those with typed questions — defined options in, one answer with a probability per option out — served by [TypeSafe Jev](https://docs.typesafe.ai) where you have access, and by a labelled Claude-backed adapter where you don't. Every piece falls back honestly; nothing breaks without a key. **[The full map, all 20 pieces →](docs/JEV-DECISION-LAYER.md)**
-
-```bash
-npm run route -- "my landlord kept my deposit"
-# security-deposit-recovery · pack pm-renters · tier high-stakes · confidence 0.83 · also: lease-decoder
-```
-
-| | |
-|---|---|
-| **Route** | a Claude Code [hook](hooks/suggest-skill-jev.sh) · `POST /route` on the [worker](mcp-remote/) · the [`pm-skills-jev-picker`](integrations/jev/) npm package · the **[Skill Router](https://mohitagw15856.github.io/pm-claude-skills/router.html)** page: paste your week's prompts, see the skills you should have used |
-| **Guard** | [input guard](integrations/jev/guard.mjs) on the free runs · [crisis router](integrations/jev/crisis.mjs) for public bots · [risk-tier second opinion](scripts/classify-risk-tiers.mjs) · [human-review queue](docs/HUMAN-REVIEW-QUEUE.md) ranked by harm |
-| **Judge** | a [typed eval judge](evals/jev-judge.mjs) · [sycophancy scan](skillbench/SYCOPHANCY.md) · [route-bench](skillbench/reports/route-bench.md): keyword floor 50.9% top-1 on 271 cases, the Claude adapter 62.4%, the Jev row waiting on a credential |
-| **Decide** | **[pm-decisions](plugins/pm-decisions/)** — [ship-or-slip](skills/ship-or-slip/SKILL.md) · [escalate-or-hold](skills/escalate-or-hold/SKILL.md) · [renew-or-churn-call](skills/renew-or-churn-call/SKILL.md) · [hire-or-pass](skills/hire-or-pass/SKILL.md): a state schema, defined options, thresholds, and a probability per option. A person still owns the call; the distribution is evidence |
-
-No skill depends on a vendor. The SKILL.md files describe *contracts* any model can serve; the adapters live in [`integrations/jev/`](integrations/jev/), and a CI gate keeps it that way.
+The full write-up of the three feature releases, with examples: **[docs/WHATS-NEW.md](docs/WHATS-NEW.md)**. Everything older: **[CHANGELOG.md](CHANGELOG.md)**.
 
 ## 💬 What can I ask it to do?
 
@@ -205,7 +151,7 @@ Say it in your own words. The description does the routing:
 | **Install in Claude Code** | `/plugin` → search **pm-skills** *(official Anthropic directory)* — or `npx pm-claude-skills add --agent claude` |
 | **Install in Cursor / Codex / Windsurf / Cline…** | `npx pm-claude-skills add --agent cursor` *(or `codex`, `windsurf`, `aider`, `cline`, `zed`…)* |
 | **Use one skill in ChatGPT / Gemini** | copy from [`exports/chatgpt/`](exports/chatgpt/) or [`exports/gemini/`](exports/gemini/) and paste as instructions |
-| **Skills over MCP, in any session** | `claude mcp add pm-skills -- npx -y pm-claude-skills-mcp` |
+| **Skills over MCP, in any session** | `claude mcp add pm-skills -- npx -y -p pm-claude-skills pm-claude-skills-mcp` |
 | **Route a prompt to a skill** | `npm run route -- "board meeting on Thursday"` |
 
 No `npm install` needed — `npx pm-claude-skills …` always runs the latest. Per-tool instructions: **[docs/installation.md](docs/installation.md)**.
@@ -374,9 +320,16 @@ Checks frontmatter, the `Use when …` trigger clause a model actually matches o
 **Companion tools** for the bits a skill shouldn't guess — both MIT, zero-dependency, no model call: **[notugly](https://github.com/mohitagw15856/notugly)** (contrast and design-system checks the design skills call instead of estimating) and **[rulebook](https://github.com/mohitagw15856/rulebook)** (37 games, 203 rulings, for the board-game skills).
 </details>
 
-## 🆕 Latest
+## 🚫 What PM Skills is not
 
-**[v80.1.0](https://github.com/mohitagw15856/pm-claude-skills/releases/latest)**: design taste, a five-stage UI pipeline (direction, system, motion, critique, verification) with pointers to five specialist design skills. v80.0.0 added the promote loop and v79 the decision layer. Everything else is in the **[changelog](CHANGELOG.md)**.
+| | |
+|---|---|
+| **Not an agent framework.** | Skills are instructions your existing assistant reads. There is no runtime to host. |
+| **Not a prompt pack.** | Every skill has required inputs, an output format, quality checks and anti-patterns, and is gated in CI. |
+| **Not only for product managers.** | PM stands for Professional. Most of the library is life admin and 35 professions. |
+| **Not a substitute for a professional.** | The legal, medical and financial skills prepare you for that conversation. They do not replace it. |
+| **Not tied to one vendor.** | The same skill exports to 12 platforms and is served over MCP. |
+| **Not a hosted service.** | The playground and the MCP endpoint are optional. Everything works from files on your machine. |
 
 ## ❓ Straight answers
 
@@ -387,7 +340,7 @@ Yes — MIT, all 1176 skills, forever. The skills are markdown; there is nothing
 Not to browse, read, install, or use skills inside a tool you already have. The playground serves a few sponsor-funded free runs a day. A key only matters for optional extras: running skills from CI, or the typed decision layer (which falls back to keyword routing without one).
 </details>
 <details><summary><b>The catalog says 1176 but the folder has more. Which is it?</b></summary>
-Both. There are <b>1178</b> directories under <code>skills/</code>; <b>12</b> of them are <a href="docs/DEPRECATION.md">deprecated</a> (marked in their frontmatter, each pointing at the skill that replaced it). They stay on disk so an old install command or a bookmarked name never breaks, and they are hidden from the catalog, the playground and the headline count. 1176 is what a person can browse; 1178 is what an installer can resolve.
+Both. There are <b>1188</b> directories under <code>skills/</code>; <b>12</b> of them are <a href="docs/DEPRECATION.md">deprecated</a> (marked in their frontmatter, each pointing at the skill that replaced it). They stay on disk so an old install command or a bookmarked name never breaks, and they are hidden from the catalog, the playground and the headline count. 1176 is what a person can browse; 1178 is what an installer can resolve.
 </details>
 <details><summary><b>I'm not a product manager. Is this for me?</b></summary>
 PM stands for <i>Professional</i> here. Most of the library is decoders for leases and medical bills, salary-negotiation practice, career-moment kits, life admin, and 35 professions from teaching to veterinary. The product-management corner is just where it started.
@@ -399,6 +352,32 @@ No. Skills are inert text files in a folder; your assistant reads them when rele
 Every skill passes a structural gate (SkillSpec L3) and a security scan in CI; 208 outputs are <a href="https://mohitagw15856.github.io/pm-claude-skills/leaderboard.html">eval-scored in the open</a> (avg 4.8/5), and the <a href="skillbench/REPORT.md">benchmark report</a> publishes the negative findings too. When something is machine-translated or unscored, it's labelled.
 </details>
 
+## 🗺️ Roadmap
+
+- ✅ Exports to 12 platforms and one install command
+- ✅ MCP server, subagents and slash commands
+- ✅ Quality gates in CI: structure, security, duplicates, drift
+- ✅ The decision layer and the promote loop
+- ✅ The design taste pipeline
+- ⚪ Published eval scores for more skills
+- ⚪ A promotion round for the Production-Ready tier
+- ⚪ Merging the near-duplicate skills still marked as pending
+- ⚪ Translated skill descriptions, done at scale or retired
+
+Direction, not a contract. Details and good first issues: **[ROADMAP.md](ROADMAP.md)**.
+
+## 🔒 Privacy and telemetry
+
+Off by default. A skill is a text file and sends nothing.
+
+| What | Default | Detail |
+|---|---|---|
+| Skills, CLI, exports | Nothing is sent | Plain files on your machine |
+| Usage counter | Off | Opt in with `PM_SKILLS_TELEMETRY=1`. It sends the skill name and nothing else |
+| Playground and hosted MCP | Only if you use them | Your prompt goes to the model provider you choose |
+
+The full statement, including the handler you can read: **[docs/TELEMETRY.md](docs/TELEMETRY.md)** and **[SECURITY.md](SECURITY.md)**.
+
 ## 🤝 Contributing
 
 <p align="center">

package/announcements/current.json

@@ -0,0 +1,12 @@
+{
+  "schemaVersion": 1,
+  "announcement": {
+    "id": "mcp-command-package-form",
+    "level": "action",
+    "title": "The MCP install command changed",
+    "body": "If you registered the MCP server with the short command, it will not start. Register it again with: claude mcp add pm-skills -- npx -y -p pm-claude-skills pm-claude-skills-mcp",
+    "url": "https://github.com/mohitagw15856/pm-claude-skills/blob/main/mcp/README.md",
+    "from": "2026-09-28",
+    "until": "2026-11-30"
+  }
+}

package/bin/doctor.mjs

@@ -6,6 +6,7 @@ import { readdirSync, existsSync, readFileSync, statSync } from 'node:fs';
 import { join, dirname } from 'node:path';
 import { fileURLToPath } from 'node:url';
 import { homedir } from 'node:os';
+import { readAnnouncement, formatAnnouncement } from './lib/announce.mjs';
 
 const PKG_ROOT = dirname(dirname(fileURLToPath(import.meta.url)));
 const HOME = homedir();
@@ -55,6 +56,10 @@ export async function run() {
 
   lines.push(`\x1b[1m🩺 pm-skills doctor\x1b[0m ${dim(`— read-only checkup · library has ${bundled.size} skills`)}\n`);
 
+  // ── 0. The current announcement, if one is active (read from the package) ──
+  const notice = readAnnouncement(PKG_ROOT);
+  if (notice) lines.push(`${formatAnnouncement(notice)}\n`);
+
   // ── 1. Skills installed per agent ─────────────────────────────────────────
   lines.push('\x1b[1mSkills\x1b[0m');
   let anyAgent = false;
@@ -115,7 +120,7 @@ export async function run() {
   if (/pm-skills|pm-claude-skills/.test(mcpTxt)) lines.push(ok('pm-skills MCP server registered with Claude Code'));
   else {
     lines.push(warn('pm-skills MCP server not registered — agents can\'t search/fetch skills on demand'));
-    rx.push([1, 'Register the MCP server: \x1b[1mclaude mcp add pm-skills -- npx -y pm-claude-skills-mcp\x1b[0m (or use the hosted URL from the README in any MCP client)']);
+    rx.push([1, 'Register the MCP server: \x1b[1mclaude mcp add pm-skills -- npx -y -p pm-claude-skills pm-claude-skills-mcp\x1b[0m (or use the hosted URL from the README in any MCP client)']);
   }
 
   // ── 4. The brain & workspace (current directory) ──────────────────────────

package/bin/lib/announce.mjs

@@ -0,0 +1,65 @@
+// The one current announcement, read from a file that ships with the package.
+// No network: `npx pm-claude-skills` always runs the latest package, so the
+// bundled file is already current. Used by `doctor` and validated by
+// scripts/check-announcement.mjs. The playground reads a copy of the same file.
+import { readFileSync, existsSync } from 'node:fs';
+import { join } from 'node:path';
+
+export const LEVELS = ['info', 'change', 'action'];
+export const LIMITS = { title: 80, body: 280 };
+const ID = /^[a-z0-9]+(?:-[a-z0-9]+)*$/;
+const DAY = /^\d{4}-\d{2}-\d{2}$/;
+
+// Returns a list of problems. An empty list means the document is valid.
+export function validate(doc) {
+  const problems = [];
+  if (!doc || typeof doc !== 'object' || Array.isArray(doc)) return ['the file must hold a JSON object'];
+  if (doc.schemaVersion !== 1) problems.push('schemaVersion must be 1');
+  if (!('announcement' in doc)) return [...problems, 'the "announcement" key is required (use null for none)'];
+  const a = doc.announcement;
+  if (a === null) return problems;
+  if (typeof a !== 'object' || Array.isArray(a)) return [...problems, 'announcement must be an object or null'];
+
+  if (typeof a.id !== 'string' || !ID.test(a.id)) problems.push('id must be kebab-case');
+  if (!LEVELS.includes(a.level)) problems.push(`level must be one of: ${LEVELS.join(', ')}`);
+  for (const field of ['title', 'body']) {
+    const v = a[field];
+    if (typeof v !== 'string' || v.trim() === '') problems.push(`${field} is required`);
+    else if (v.length > LIMITS[field]) problems.push(`${field} is ${v.length} characters, the limit is ${LIMITS[field]}`);
+    else if (/[<>]/.test(v)) problems.push(`${field} must be plain text, no markup`);
+  }
+  if (a.url !== undefined && (typeof a.url !== 'string' || !/^https:\/\/[^\s<>"']+$/.test(a.url))) problems.push('url must be an https address');
+  for (const field of ['from', 'until']) {
+    const v = a[field];
+    if (typeof v !== 'string' || !DAY.test(v) || Number.isNaN(Date.parse(`${v}T00:00:00Z`))) problems.push(`${field} must be a date as YYYY-MM-DD`);
+  }
+  if (problems.length === 0 && a.until < a.from) problems.push('until is before from');
+  return problems;
+}
+
+// `today` is a YYYY-MM-DD string so the comparison is a plain string compare.
+export function isActive(a, today) {
+  return Boolean(a) && a.from <= today && today <= a.until;
+}
+
+export function activeAnnouncement(doc, today = new Date().toISOString().slice(0, 10)) {
+  if (validate(doc).length) return null;
+  return isActive(doc.announcement, today) ? doc.announcement : null;
+}
+
+export function readAnnouncement(root, today) {
+  const p = join(root, 'announcements', 'current.json');
+  if (!existsSync(p)) return null;
+  try {
+    return activeAnnouncement(JSON.parse(readFileSync(p, 'utf8')), today);
+  } catch {
+    return null;
+  }
+}
+
+export function formatAnnouncement(a) {
+  const label = { info: 'Notice', change: 'Changed', action: 'Action needed' }[a.level] || 'Notice';
+  const lines = [`\x1b[1m📣 ${label}: ${a.title}\x1b[0m`, `  ${a.body}`];
+  if (a.url) lines.push(`  \x1b[2m${a.url}\x1b[0m`);
+  return lines.join('\n');
+}

package/bin/postinstall.mjs

@@ -45,7 +45,7 @@ try {
     `  ${A('https://mohitagw15856.github.io/pm-claude-skills/')}`,
     '',
     `  ${B('Want them in every Claude session?')} Add the MCP server:`,
-    `  ${B('claude mcp add pm-skills -- npx -y pm-claude-skills-mcp')}`,
+    `  ${B('claude mcp add pm-skills -- npx -y -p pm-claude-skills pm-claude-skills-mcp')}`,
     '',
     D('  Tip: you can skip `npm install` entirely — `npx pm-claude-skills …` always runs the latest.'),
     '',

package/connectors/README.md

@@ -11,7 +11,7 @@ you copy-pasting context.
 
 **Claude Code** — one line each:
 ```bash
-claude mcp add pm-skills   -- npx -y pm-claude-skills-mcp
+claude mcp add pm-skills   -- npx -y -p pm-claude-skills pm-claude-skills-mcp
 claude mcp add filesystem  -- npx -y @modelcontextprotocol/server-filesystem /path/to/your/project
 claude mcp add github      -- npx -y @modelcontextprotocol/server-github      # set GITHUB_PERSONAL_ACCESS_TOKEN
 ```

package/connectors/composio.md

@@ -10,7 +10,7 @@ Pair `pm-skills` (the *structure* layer) with Composio (the *action* layer) and
 
 1. **Add the `pm-skills` server** (structure layer — no auth):
    ```bash
-   claude mcp add pm-skills -- npx -y pm-claude-skills-mcp
+   claude mcp add pm-skills -- npx -y -p pm-claude-skills pm-claude-skills-mcp
    ```
 2. **Add Composio** (action layer) and authorize the apps you want. Composio handles each app's OAuth for you — you connect Gmail/Slack/Jira/etc. **once** in Composio, then the assistant can act through it. See [Composio's MCP docs](https://docs.composio.dev) for the current `mcp add` command and your hosted server URL; it looks like:
    ```bash

package/connectors/contentgoldmine.md

@@ -25,7 +25,7 @@ PM Skills is the **craft layer** for content creators: skills that teach an AI *
 ContentGoldMine's per-platform transformers are exactly the kind of structured instructions these skills encode. Instead of hand-maintaining prompts, you can have it pull them from this library's MCP server:
 
 ```bash
-claude mcp add pm-skills -- npx -y pm-claude-skills-mcp
+claude mcp add pm-skills -- npx -y -p pm-claude-skills pm-claude-skills-mcp
 ```
 
 Then a transformer fetches the relevant skill (`get_skill content-repurposer`, `get_skill short-form-script`, `get_skill newsletter-writer`) and applies it — so the app and the open library stay in sync. See [`../mcp/README.md`](../mcp/README.md).

package/connectors/mcp-pairings.md

@@ -45,7 +45,7 @@ The skills are reachable from every agent convention in the wild — pick whiche
 
 | Protocol | How |
 |---|---|
-| **MCP** (Claude Desktop/Code, Cursor, ChatGPT…) | `claude mcp add pm-skills -- npx -y pm-claude-skills-mcp`, or the hosted URL as a connector. Includes `run_skill` via **MCP sampling** (zero API key). |
+| **MCP** (Claude Desktop/Code, Cursor, ChatGPT…) | `claude mcp add pm-skills -- npx -y -p pm-claude-skills pm-claude-skills-mcp`, or the hosted URL as a connector. Includes `run_skill` via **MCP sampling** (zero API key). |
 | **A2A** (Google agent ecosystem) | Discovery card at `https://pm-skills-mcp.pm-claude-skills.workers.dev/.well-known/agent-card.json`; JSON-RPC `message/send` at `/a2a`. |
 | **AGENTS.md** (Codex, Jules, et al.) | `npx pm-claude-skills init` writes an AGENTS.md wiring the brain, the skills, and the arena artifacts for any agent that reads the convention. |
 | **Function calling** (OpenAI SDK, Vercel AI SDK, LangChain) | `npm i pm-skills-tools` — the library as typed tools with pick()/search(). |

package/connectors/mcp.example.json

@@ -3,7 +3,7 @@
   "mcpServers": {
     "pm-skills": {
       "command": "npx",
-      "args": ["-y", "pm-claude-skills-mcp"]
+      "args": ["-y", "-p", "pm-claude-skills", "pm-claude-skills-mcp"]
     },
     "filesystem": {
       "command": "npx",

package/mcp/README.md

@@ -5,7 +5,7 @@ A zero-dependency [Model Context Protocol](https://modelcontextprotocol.io) serv
 ## One-line install (Claude Code)
 
 ```bash
-claude mcp add pm-skills -- npx -y pm-claude-skills-mcp
+claude mcp add pm-skills -- npx -y -p pm-claude-skills pm-claude-skills-mcp
 ```
 
 That's it — all 174 skills and 5 workflow recipes are now available in every Claude Code session, on any project.
@@ -31,7 +31,7 @@ That's it — all 174 skills and 5 workflow recipes are now available in every C
   "mcpServers": {
     "pm-claude-skills": {
       "command": "npx",
-      "args": ["-y", "pm-claude-skills-mcp"]
+      "args": ["-y", "-p", "pm-claude-skills", "pm-claude-skills-mcp"]
     }
   }
 }

package/mcp/server.mjs

@@ -2,9 +2,9 @@
 // pm-claude-skills MCP server — exposes the skill library to any MCP client
 // (Claude Desktop, etc.) over stdio. Tools: list_skills, search_skills, get_skill.
 //
-// Run directly: node mcp/server.mjs   (or, once published: npx pm-claude-skills-mcp)
+// Run directly: node mcp/server.mjs   (or, once published: npx -y -p pm-claude-skills pm-claude-skills-mcp)
 // Configure in an MCP client, e.g. Claude Desktop claude_desktop_config.json:
-//   { "mcpServers": { "pm-claude-skills": { "command": "npx", "args": ["-y", "pm-claude-skills-mcp"] } } }
+//   { "mcpServers": { "pm-claude-skills": { "command": "npx", "args": ["-y", "-p", "pm-claude-skills", "pm-claude-skills-mcp"] } } }
 //
 // Pure Node standard library — no dependencies. Protocol: newline-delimited
 // JSON-RPC 2.0 (the MCP stdio transport). All logging goes to stderr so it

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "pm-claude-skills",
-  "version": "80.1.0",
+  "version": "80.2.0",
   "type": "module",
   "mcpName": "io.github.mohitagw15856/pm-claude-skills",
   "description": "1176 professional Agent Skills (SKILL.md) + subagents + slash commands for Claude, ChatGPT, Gemini, Cursor, Codex & Hermes. It's a CLI, not a library: run `npx pm-claude-skills add --agent <tool>` to install into your AI tool — no `npm install` needed. Or try any skill free at https://mohitagw15856.github.io/pm-claude-skills/",
@@ -51,6 +51,7 @@
     "exports/",
     "templates/",
     "connectors/",
+    "announcements/",
     "skill-tiers.json",
     "skill-sources.json",
     "workflows.json",
@@ -64,6 +65,7 @@
     "new-bundle": "node scripts/new-bundle.mjs",
     "skillcheck": "node scripts/skillcheck.mjs",
     "skill-dupes": "node scripts/skill-dupes.mjs",
+    "skill-release": "node scripts/skill-release.mjs",
     "build:disambiguation": "node scripts/build-disambiguation.mjs",
     "build:exports": "node scripts/build-exports.mjs",
     "build:web": "node web/build-skills.mjs",
@@ -82,18 +84,23 @@
     "improve": "node scripts/improve-skill.mjs",
     "eval:gen-cases": "node scripts/gen-eval-cases.mjs",
     "eval:all": "node scripts/gen-eval-cases.mjs && node evals/run-evals.mjs --cases evals/cases.full.json --models claude-haiku-4-5-20251001 --judge claude-haiku-4-5-20251001",
+    "evals:promptfoo": "cd evals/promptfoo && PROMPTFOO_DISABLE_TELEMETRY=1 npx -y promptfoo@0.123.1 eval -c promptfooconfig.yaml",
+    "evals:promptfoo:smoke": "cd evals/promptfoo && PROMPTFOO_DISABLE_TELEMETRY=1 npx -y promptfoo@0.123.1 eval -c promptfooconfig.smoke.yaml --no-cache --no-write --no-progress-bar --no-table",
     "check:drift": "node scripts/check-drift.mjs",
     "lint:exports": "node tests/exports-lint.mjs",
     "check:weight": "node scripts/check-web-weight.mjs --check",
+    "check:design": "node scripts/check-design-tokens.mjs",
     "test:scripts": "python3 -m unittest discover -s tests -p 'test_*.py'",
     "test:skillcheck": "node tests/skillcheck-test.mjs",
     "check:conflicts": "node scripts/check-antipattern-conflicts.mjs",
     "build:badge": "node scripts/build-conformance-badge.mjs",
     "check:a11y": "node scripts/check-a11y.mjs --check",
     "check:evals": "node scripts/check-eval-coverage.mjs --check",
+    "check:announcement": "node scripts/check-announcement.mjs --check",
     "tier-review": "node scripts/tier-review.mjs",
     "staleness": "node scripts/check-staleness.mjs",
     "route": "node integrations/jev/route.mjs",
+    "verify:release": "node scripts/release-verify.mjs",
     "check:jev": "node integrations/jev/client.mjs --selftest && node integrations/jev/adapter.mjs --selftest && node integrations/jev/catalog.mjs --selftest && node integrations/jev/route.mjs --selftest && node integrations/jev/suggest.mjs --selftest && node integrations/jev/guard.mjs --selftest && node integrations/jev/crisis.mjs --selftest && node integrations/jev/decide.mjs --selftest && node scripts/journey-gate.mjs --selftest && node scripts/classify-risk-tiers.mjs --selftest && node scripts/check-vendor-neutrality-semantic.mjs --selftest && node scripts/human-review-queue.mjs --selftest && node evals/jev-judge.mjs --selftest && node scripts/sycophancy-scan.mjs --selftest && node scripts/drift-triage.mjs --selftest && node scripts/example-output-gate.mjs --selftest && node skillbench/route-bench.mjs --selftest && node scripts/skill-of-the-week-jev.mjs --selftest && node extension/jev-suggest.js --selftest"
   },
   "devDependencies": {

package/skills/competitive-analysis/SKILL.md

@@ -1,6 +1,7 @@
 ---
 name: competitive-analysis
 description: "Analyze competitors and create competitive landscape documentation with feature matrices, positioning maps, and strategic recommendations. Use when asked to analyze competitors, create competitive analysis, compare features with competitors, build a competitive landscape, track competitive positioning, or prepare sales battlecard inputs. Produces structured competitor profiles, feature comparison matrix, win/loss analysis, and prioritised strategic recommendations. For a one-off teardown of a single rival use competitor-teardown; for a recurring market briefing use competitive-intelligence-monitor."
+version: 1.0.0
 ---
 
 # Competitive Analysis Skill

package/skills/meeting-notes/SKILL.md

@@ -1,6 +1,7 @@
 ---
 name: meeting-notes
 description: "Structure and format meeting notes following PM best practices. Use when asked to create meeting notes, format discussion notes, capture action items, or document decisions from any meeting type. Produces structured notes with decisions, action items (owner + deadline), open questions, and next steps."
+version: 1.0.0
 ---
 
 # Meeting Notes Skill

package/skills/prd-template/SKILL.md

@@ -1,6 +1,7 @@
 ---
 name: prd-template
 description: "Create a Product Requirements Document following proven PM template structure. Use when asked to write a PRD, product spec, feature specification, or requirements document for a new feature or product. Produces a complete PRD with problem statement, user stories, functional requirements, technical considerations, and success metrics."
+version: 1.0.0
 ---
 
 # PRD Template Skill

package/skills/stakeholder-update/SKILL.md

@@ -1,6 +1,7 @@
 ---
 name: stakeholder-update
 description: "Create concise executive stakeholder updates using the BLUF (Bottom Line Up Front) framework. Use when asked to write a status update, progress report, project communication, or executive briefing for leadership or stakeholders. Produces a BLUF-led update with status, key metrics, risks, upcoming milestones, and decisions needed — readable in under 2 minutes."
+version: 1.0.0
 ---
 
 # Stakeholder Update Skill

package/skills/user-research-synthesis/SKILL.md

@@ -1,6 +1,7 @@
 ---
 name: user-research-synthesis
 description: "Analyze and synthesize user research findings into structured, actionable insights. Use when given user research data, interview transcripts, survey results, or user feedback that needs to be analyzed and summarised. Produces a themed synthesis with prevalence data, supporting quotes, pain points analysis, feature request prioritisation, and recommended next steps. For interview transcripts specifically use user-interview-synthesis instead."
+version: 1.0.0
 ---
 
 # User Research Synthesis Skill

package/templates/apprentice/package.json

@@ -5,6 +5,6 @@
   "description": "A ready-to-run Claude Agent SDK agent grounded in your pm-skills workspace: your brain, your skills, your standards.",
   "type": "module",
   "scripts": { "start": "node index.mjs" },
-  "dependencies": { "@anthropic-ai/claude-agent-sdk": "^0.1.0" },
+  "dependencies": { "@anthropic-ai/claude-agent-sdk": "^0.3.282" },
   "engines": { "node": ">=18" }
 }