pi-skills

A set of everyday agent skills: web search, browser automation, Gmail, Google Calendar and Drive, speech transcription and YouTube transcripts

  • 2,657 ★ on the repository
  • Last commit
  • MIT
Skill

Extended access

Why this level

  • Runs local CLIs and scripts and reaches the internet through external APIs
  • Acts on personal mail, calendar and drive: creates events, drafts and edits files
  • Automates Chrome and executes scripts on pages
All reasons and checks

badlogic/pi-skills

Install

In your terminal, with SkillFoxx CLI

npx skillfoxx add skills/pi-skills

Detects the agents on your machine, shows the access level and pins the version.

Other ways to install

Access and checks

How we review

Extended accessThe tool runs code or makes network calls and only reads external data.

  • Runs local CLIs and scripts and reaches the internet through external APIs
  • Acts on personal mail, calendar and drive: creates events, drafts and edits files
  • Automates Chrome and executes scripts on pages

This is third-party code. Review the repository files before installing.

What it does

The set adds several self-contained skills for everyday tasks. brave-search provides web search and page content extraction via the official Brave API without a browser. browser-tools drives Chrome over the DevTools Protocol: navigation, screenshots, element picking and script execution. gccli, gdcli and gmcli are command-line wrappers over Google Calendar, Drive and Gmail for viewing and creating events, managing files and handling mail. transcribe turns speech into text via Whisper on Groq, youtube-transcript fetches video transcripts, and vscode helps compare files and show diffs. Each skill is a SKILL.md with instructions and helper files, and the agent picks the right one from its description. The set is built for pi-coding-agent and is compatible with Claude Code, Codex CLI, Amp and Droid.

Who it is for. For developers who want ready skills for search, the browser and Google services straight from the agent.

Good fit when

  • You need web search or page content extraction without an interactive browser
  • You need Chrome automation: navigation, screenshots, running scripts
  • You need Gmail, Google Calendar and Drive from the command line

Not a fit when

  • You have no keys or accounts for the needed services: Brave, Google, Groq
  • Policy forbids connecting the agent to personal mail, calendar and drive
  • You want a single MCP server rather than a set of separate CLIs and skills

Example request

Search the web for current docs on this library and extract the content of the first result

Limitations

The skills are thin wrappers over external services, so keys and accounts are required: a Brave Search account with a free subscription (a card is required to create it, with no charges), a Google Cloud project with OAuth for calendar, drive and mail, and a Groq key for transcription. Some skills need Node.js and npm packages, browser-tools needs Chrome, and vscode needs the code CLI in PATH. The Google wrappers act on the user's personal data: they create events, drafts and manage files, so access rights need care.

How to disable. Remove the symlinks or skill folders from ~/.claude/skills/ or the agent's directory (.pi, .codex, .factory). Globally installed CLIs can be removed with npm uninstall -g.

FAQ

Is this one skill or a set?

A set of separate skills: web search, browser, Gmail, calendar, drive, speech transcription, YouTube transcripts and VS Code integration. Install only what you need.

Does it only work with pi?

No. The set is built for pi-coding-agent but is compatible with Claude Code, Codex CLI, Amp and Droid, each with its own install path in the README.

Discussion

No questions or reviews yet.

Skills

cafe3310's personal skill collection: spec-driven project work, reviews, reports, browser, 3D and video in one repository

SkillExtended access256Repository stars

Twenty-two standalone skills: Notion, Todoist, Tavily, Context7, image generation, converting documents to Markdown

SkillBroad access302Repository stars

18 portable skills for work where AI is weakest: messy voice input, message tone, small trend tests and safe shipping.

PluginMinimal access108Repository stars

A personal set of six skills: goal briefs for an agent, disk cleanup, AI news digests, project tidy-up, deep research and long-form writing

SkillExtended access22.2KRepository stars

A library of 229 skills for work and life: sales, marketing, finance, engineering, design and parallel agent runs

SkillBroad access428Repository stars

A personal library of 25 skills: posters and storyboards, Google-style code review, lab report reading and scenario simulations

SkillExtended access629Repository stars

README in short

The README lists eight skills and gives install paths for pi-coding-agent, Codex CLI, Amp, Droid and Claude Code. In Claude Code, which looks for SKILL.md one level deep, each skill folder is attached by its own symlink. It documents the skill format: YAML frontmatter with name and description, instructions and a {baseDir} placeholder replaced with the skill's directory path at runtime. The requirements section notes which skills need Node.js, npm packages, Chrome, Brave and Groq keys or VS Code. MIT licensed.

README badge

Original SKILL.md

---
name: brave-search
description: Web search and content extraction via Brave Search API. Use for searching documentation, facts, or any web content. Lightweight, no browser required.
---

# Brave Search

Web search and content extraction using the official Brave Search API. No browser required.

## Setup

Requires a Brave Search API account with a free subscription. A credit card is required to create the free subscription (you won't be charged).

1. Create an account at https://api-dashboard.search.brave.com/register
2. Create a Free AI subscription
3. Create an API key for the subscription
4. Add BRAVE_API_KEY to your shell profile
5. Install dependencies (run once): cd {baseDir} && npm install

## Search

{baseDir}/search.js "query"                 # Basic search (5 results)
{baseDir}/search.js "query" -n 10           # More results (max 20)
{baseDir}/search.js "query" --content       # Include page content as markdown

Report a problem