Browse agent skills

Capability SDK with all tools, structured responses, AI SDK/MCP adapters and automatic Windows/Linux runtime setup.

A zero-dependency, universal HTTP client built on the Fetch API. Works in browsers, Node 20+, Deno, Bun, and Cloudflare Workers.

Client-side PDF compression: extract & recompress embedded images with hardware-accelerated native Canvas / libjpeg-turbo + pdf-lib rebuild. No server, no UI. Ships as raw ESM source for Vite-based apps (React + Vite, Astro).

Gu CLI — the gu terminal plus the local worker that runs agent / OCR / PDF / embedding jobs on your Mac (Ollama / MLX / OpenAI-compatible).

AWS services like S3, DynamoDB, and Lambda inside Node or a browser tab, reached with the ordinary AWS SDK

Electron host adapter and plugin for open-snapora

VS Code extension to notify you when an agent or workspace needs attention

Run Ollama models with full tool and vision support in GitHub Copilot Chat. Manage local and remote Ollama instances directly from VS Code.

Context-aware voice input for pi. Speak a request, get a formulated prompt at the cursor.

GPU server monitoring and AI service manager web dashboard

Minimal web search and URL fetching for Pi coding agent. Search via OpenAI or Brave; fetch readable/raw content from URLs, GitHub repos, and PDFs.

A local-first, cross-vendor multi-agent web harness for shared-context work on git projects.

Component search, schematic layout and PCB placement for EDA Copilot integrations

Timbro's Node SDK for Payments and Payment invoice coordination.

Live session status in your terminal/tab title for pi: provider glyph, working/needs-you/resting state, and what it's doing — event-driven, no LLM calls.

Paintless destination adapter that POSTs requests and results to any webhook

Z Research PPT stage — presentation generation (Beamer/Quarto) from the paper. Ships one skill.

Z Research report stage — academic paper writing for asset pricing + ML/DL research. Ships one skill.

Z Research analysis stage — end-to-end model training and estimation for asset pricing + ML/DL research. Ships one skill.

Practical guardrails for AI coding agents. Hold on, think first, code less.

Z Research data stage — data discovery and quality assessment for asset pricing + ML/DL research. Ships one skill.

Z Research literature stage — literature search, synthesis, and citation-chain tracking for asset pricing + ML/DL research. Ships one skill.

Manage a shared Codex base prompt for an OpenClaw agent.

Desktop and Android Web Push notifications for Paperclip board events.

Offline /voice dictation for pi. Capture with a bundled PvRecorder binary, transcribe with a local faster-whisper. No ffmpeg and no native build step.

Review your coding agent's changes like a pull request — locally, with inline comments and threaded follow-ups, then hand the marked-up review back to the agent

T Expert — a 323-expert, 22-division roster with full Chinese translations, as a standalone DeepSeek Harness plugin, with a built-in multi-agent team engine (T Team).

HookNiew — Webhook reliability relay (Rust + Redis + Postgres). Queued -> Delivered -> Logged

Human review before DeepSeek Harness context compaction

@Doc — Lexer/Parser/Adapters/registry for the @Doc document language, plus the Monarch editor tokenizer.