RP

romiluz13/pi-agent-skills

Developer tools
20 stars Quality 47 Trend 31

This repo packages 11 skills that answer Pi questions from a pinned pi-mono checkout instead of guessing from memory. Every cited pi-mono/... path is verified on disk.

Overview

This repo packages 11 skills that answer Pi questions from a pinned pi-mono checkout instead of guessing from memory. Every cited pi-mono/... path is verified on disk. Every shipped eval, graded baseline, and trigger set is checked in CI. If you want Pi-focused answers that are package-aware, file-aware, and harder to hallucinate, this repo is the point. - Pi documentation is split across package READMEs, docs/, examples, and source files. These skills turn that into package-level entry points. - Answers are forced to stay grounded in real pi-mono/... paths. Broken citations fail verification. - The repo ships developer-facing checks, not just Markdown: eval catalogs, graded baselines, trigger sets, and verification scripts live alongside the skills. Read docs/working-with-pi-mono.md before editing anything substantial. It explains the clone layout, corpus pinning, and where upstream Pi docs actually live. Use the repo as a skills bundle from GitHub: The repo also ships a .

README

Pi Agent Skills

Source-grounded skills for Pi.

This repo packages 11 skills that answer Pi questions from a pinned pi-mono checkout instead of guessing from memory. Every cited pi-mono/... path is verified on disk. Every shipped eval, graded baseline, and trigger set is checked in CI.

If you want Pi-focused answers that are package-aware, file-aware, and harder to hallucinate, this repo is the point.

Why use these skills

  • Pi documentation is split across package READMEs, docs/, examples, and source files. These skills turn that into package-level entry points.
  • Answers are forced to stay grounded in real pi-mono/... paths. Broken citations fail verification.
  • The repo ships developer-facing checks, not just Markdown: eval catalogs, graded baselines, trigger sets, and verification scripts live alongside the skills.

Coverage

Skill Focus
pi-cli-workspace CLI behavior, settings, sessions, compaction, /tree, providers, models, philosophy
pi-extension-authoring ExtensionAPI, commands, tools, dynamic resources, TUI integration, custom providers
pi-package-authoring Pi packages, manifests, install/update behavior, precedence
pi-ai-library @mariozechner/pi-ai, providers, streaming, tool calling, provider work
pi-agent-embedding @mariozechner/pi-agent-core, events, embedding, SDK examples
pi-web-ui @mariozechner/pi-web-ui, storage, example app, browser integration
pi-rpc-sdk RPC mode, JSON mode, AgentSession, SDK entry points
pi-tui @mariozechner/pi-tui, components, overlays, rendering, input handling
pi-mom Slack bot, sandboxing, events, artifacts server, multi-platform docs
pi-pods GPU pod setup, vLLM config, model startup, tool parsers
pi-customization themes, keybindings, prompt templates, system prompt overrides

Repo Layout

Path Purpose
pi-skills/ Canonical skills source: SKILL.md, references, evals, trigger sets, scripts, tests
skills/ Symlinked install surface for the open skills CLI
pi-mono/ Local clone used as the source corpus when developing or verifying this repo
docs/ Repo docs for corpus layout, grading needles, and related workflows
.claude-plugin/ Claude Code plugin manifest for loading the repo as a plugin

Read docs/working-with-pi-mono.md before editing anything substantial. It explains the clone layout, corpus pinning, and where upstream Pi docs actually live.

Install

Open Skills CLI

Use the repo as a skills bundle from GitHub:

# Install all skills
npx skills add romiluz13/pi-agent-skills -a pi -y

# Inspect what the bundle exports
npx skills add romiluz13/pi-agent-skills --list

# Install a single skill
npx skills add https://github.com/romiluz13/pi-agent-skills/tree/main/skills/pi-cli-workspace -a pi -y

Claude Code Plugin

The repo also ships a .claude-plugin/plugin.json manifest.

claude --plugin-dir /path/to/pi-agent-skills

That exposes the skills under the pi-skill: namespace, for example:

/pi-skill:pi-cli-workspace
/pi-skill:pi-extension-authoring
/pi-skill:pi-ai-library

Manual Pi Install

Pi can also load the skills directly from package.json pi.skills, or from copied / symlinked skill directories under ~/.pi/agent/skills/.

Optional Cursor symlinks:

SKILLS="$PWD/pi-skills"
for d in pi-cli-workspace pi-extension-authoring pi-package-authoring pi-ai-library pi-agent-embedding pi-web-ui pi-rpc-sdk pi-tui pi-mom pi-pods pi-customization; do
  ln -sf "$SKILLS/$d" "$HOME/.cursor/skills/$d"
done

Verify

cd pi-skills && npm run verify-skills
cd pi-skills && npm run ci

verify-skills runs the static repo gates: corpus paths, corpus pin, frontmatter, eval shape, pack sync, graded-needle derivation, graded-response checks, trigger-set validation, and more.

npm run ci adds the checked-in unit tests and regrades all baselines before running the same verification stack.

When you change eval definitions:

cd pi-skills
npm run sync-evals
npm run grade-graded-examples

Local Development

This repo does not vendor pi-mono in Git. For local development, clone pi-mono into ./pi-mono next to pi-skills/.

git clone https://github.com/romiluz13/pi-agent-skills.git
cd pi-agent-skills
git clone https://github.com/badlogic/pi-mono.git pi-mono
cd pi-mono && npm install && npm run build && cd ..
cd pi-skills && npm run verify-skills

When you intentionally move the corpus pin:

cd pi-skills
npm run bump-corpus
npm run verify-skills
npm run sync-evals

Why this repo is useful to developers

This repo is not trying to be a second set of docs for Pi.

It is a developer toolchain for trustworthy answers about Pi:

  • package-specific entry points instead of “search the monorepo and hope”
  • skills that can explain behavior from source, docs, and examples together
  • machine-checked evals that keep those skills honest over time

If you already work in Pi, that means less spelunking and fewer wrong answers. If you maintain Pi-related tooling, it means you can reuse the same grounded corpus in Pi, Claude Code, or any workflow that can consume skills.

Further Reading

View the full README on GitHub

Recommended Tools

Try a different keyword or remove a filter.

Install

npx skillfish add romiluz13/pi-agent-skills
New to MCP? Start with the mcpskill.org guide →