Loading...
Directory
Search by intent: design systems, testing, deployment, security, documentation, Azure, Supabase, React, or any workflow you want an agent to perform better.
Showing 18,031–18,060 of 127,302 skills
Page 602 of 4,244 · loading 30 at a time
Route and read documents into clean Markdown — Word, Excel, PowerPoint, OpenDocument, RTF, EPUB, CSV and text-layer PDF. Names the pages it could not read instead of dropping them silently. Text recognition is not included: it needs a native runtime.
对话式产品需求讨论 Agent Skill（Claude Code / Codex CLI 通用）：需求先于方案、竞品现实、结构化反方，把模糊想法推进到 PRD 或想清楚的「不做」/ An adversarial, evidence-first product-requirements discussion skill for SKILL.md-compatible agents
DeepSeek V4 on the Responses API with native server-side web search (OpenCode Zen/Go + DeepSeek first-party) — search markers shown in the thinking chain
Zero-dependency parser for HTTP Accept-Encoding request headers with q-value and parameter support (RFC 9110 §12.5.3)
LangChain Messages provider for @launchdarkly/ai-server
Claude (Anthropic OAuth/subscription) compatibility layer for pi-coding-agent — keeps ALL extension tools usable under the Claude subscription by mapping unknown flat tool names to mcp__pi__<name> on the wire instead of dropping them.
Delegate coding-agent work by spawning other agents as herdr tabs.
Agent Discussion Arena per gsd-pi: consiglio di agenti con ruoli configurabili, coordinato dal ciclo auto di gsd-pi
Cohere Command A+ vocabulary for hypertok format version 1
Standalone read-only dashboard for Playwright Backend Mocks
Meta Model API OAuth provider and Muse-style voice dictation for pi
Turn any repo into a self-organizing AI engineering team — one command, 8 specialists, evidence-gated delivery. For opencode.
Node.js / Next.js client for fetching content from Draftbase
The public client for the Bowmark capability library — real TypeScript for the whole bowmark.* surface, with no Bowmark source on the caller's disk. ZERO runtime dependencies, deliberately and permanently.
Text-to-speech plugin for OpenCode — hear LLM responses spoken aloud using Kokoro or Supertonic 3
The pure annotation brain: model · update(msg)→[model,effects] · view. Content-space geometry per family, hit-testing, and tools. No DOM, engine, or framework dependencies.
CLI tool to scaffold modern Express.js applications with TypeScript, ORMs, and best practices
Low-level PDF execution runtime for EmbedPDF, backed by EmbedPDF's fork of PDFium and distributed as WASM and native platform packages.
Pi package for Diffbot — structured web knowledge tools for the Pi coding agent
Pi extension that classifies every final assistant reply against markdown output rules and asks the model to rewrite violations. Ships with an ASD-STE100 Simplified Technical English + TLDR rule.
LiveKit adapters for the Glove voice stack — room transport for realtime agents, and avatars that join the room as participants
Live avatar adapters for the Glove agent framework — a face over the speech-to-speech voice stack
Record once. Replay forever. Deterministic tests for LLM apps — record real LLM API interactions, replay them in CI: zero tokens, zero flakiness.
Apple Intelligence for Tauri: guest bindings for tauri-plugin-apple-intelligence plus a Vercel AI SDK provider with pluggable transports
Provision and manage an agent-operated Zoneless store
n8n community nodes for Gluecrawl — agentic web scraping. Create jobs from a plain-English goal, start runs, pull structured rows, and trigger workflows on run completion.
n8n community node for the Bluesky API
Apache-2.0 CLI for AgentVM managed cloud desktops, servers, retail usage billing, and OS access.