Browse agent skills

Deterministic DAG execution runtime with checkpointing, retries, state persistence, and incremental execution support.

A small YAML DSL for describing application feature logic — validate it, visualize it with Mermaid, and use it as a behavioral source of truth for humans and AI coding agents.

Schema-defined document graphs with hybrid search, graph retrieval, and composable storage

A fast, distraction-free torrent search engine and downloader built for your terminal.

Cursor stays the brain. Grok CLI + DeepSeek (claude-ds) + OpenRouter easy lane are the parallel army — lane-aware /route orchestration in tmux.

Plugin Paperclip de observabilidade operacional via Telegram (somente leitura).

Deterministic application planning engine for validating, normalizing, and transforming application manifests into executable DAG-based execution plans.

GitHub provisioning tools for the TAI ecosystem: manifest-loaded repository-webhook setup tools that create a repository webhook, list a repository's webhooks with Link-header pagination, and delete a repository webhook — the provider-side counterpart to the runtime signature verifier.

Twilio provisioning tools for the TAI ecosystem: manifest-loaded REST tools that list a Twilio account's incoming phone numbers, read a number's inbound-message webhook, and set a number's inbound-message webhook URL.

Turn website references into original, product-specific design systems for ChatGPT and Codex.

Incremental LLM tool-call JSON parser and deterministic repair engine — never reports a value until it's unambiguously complete

Proactive semantic memory and mid-run steering for the Pi coding agent

Shared Model for InferenceService

An autonomous AI Operating System / Framework for Vibe Coding. Installs a highly optimized 5-Pillar context brain into your project.

Image understanding for text-only LLMs in pi — paste transcription + read_image tool via your configured vision providers

OpenContext · rag module

A high-performance, developer-friendly, fully typed TypeScript/JavaScript SDK for interacting with the Skyrme Chat V2 and V3 Enterprise APIs.

Local collector for LLM token usage emitted through OpenTelemetry

ICAO 9303 Iraqi National ID OCR + NFC reader for React Native / Expo

Installs the Product Atlas skill into your repo for Claude Code.

Zero-shot decision tree induction and embedding using large language models

生成面试题+考察维度+A级参考答案的完整面试准备系统 | Interview Question Generator with Assessment Dimensions and Reference Answers

Zero-friction LLM footprint reporting for saive (saive.tech) — patches OpenAI, Anthropic and Google GenAI clients to report usage asynchronously. Never sends prompt or completion content.

n8n community nodes for the Grok API from xAI — text with reasoning (thinking) mode, image generation & editing, video generation, and voice (text-to-speech & speech-to-text). OpenAI-compatible, with a configurable Base URL.

A professional n8n community node for OpenRouter and any OpenAI-compatible API: chat completions with reasoning/thinking mode, model fallbacks, provider routing, tool calling and more.

A thin, stacked context-window composition bar for Pi.

Type-safe GraphQL support for the commercetools TypeScript SDK, with a schema derived query builder.

LaTeX class, reference material and format checker for UMaT (Tarkwa) theses and project reports — standalone, or as a Claude Code skill.

适用于 Claude Code 和 Codex 的求职 Skill：查找岗位、判断是否值得投递、制作简历与申请材料、检验申请材料并模拟面试。不编造技能或业绩数字，不预测录用概率；检查简历是否便于招聘管理系统（ATS）读取，并从招聘人员和用人经理视角分别独立评审。

Connect Claude Code, Codex and Cursor telemetry to your backend with turn, Git and ticket-key attribution.