Runtime-agnostic agent skill for evidence-bound supplier discovery reports. Works with Claude Code, Cursor, OpenCode, Codex, Gemini CLI.

Agent skills by tarte, using gh skill

Autonomous optimization-loop skill for Claude (try, measure, keep, revert, repeat). A Claude reimplementation of pi-autoresearch. Install via: npx skills add agusmdev/autoresearch-skill

Agent Skill for reviewing AI-generated pull requests — triage, red flags, CI gaming, duplication, and workflow security.

GPU-native screen vision for AI agents — DXGI Desktop Duplication capture + compositor dirty-rect change feed. 96% less image data than full-frame screenshots.

Agent skill for AI inference — 200+ LLMs, Bittensor, image gen, embeddings, GPU. OpenAI-compatible, pay-per-call x402, USDC, no API keys.

Visual companion workflow skill for Claude Code and Codex agents

My personal collection of agent skills for AI-powered workflows

Dumb Army — fan out an army of dumb-but-real persona subagents over real screenshots to find where users get confused (burro-first UX/copy testing)

Agent skills to decide, generate, and set up frontend tests (unit vs e2e). Framework-agnostic, installable via npx skills add.

Local-first, agent-friendly AIOps vault template with Markdown/JSONL truth sources and thin companion skills.

蒸馏一个人的认知操作系统的 Agent Skill — 对外汉语教学法 × AI「第零阶级」方法论 × 跨时空连接思维。Distill a person's cognitive OS. Inspired by nuwa-skill.

A context-preserving Git/GitHub safety skill for AI coding agents.

Claude Code skill: pull clean Polymarket data (markets/prices/trades) with no keys, no 403s, no double-counted trades.

Practical bilingual skills that make AI coding agents more disciplined, testable, and useful in real engineering work.

Starter agents + skills for TOA Agent Build Day (June 27, 2026) — 9 fork-ready projects, agentskills.io-spec skills, measured evals, local-first

OpenAI Codex skill that creates verified visual user guides and cited developer documentation from real repositories.

Turn GitHub issues into ready-to-run task packs for coding agents.

Access app store ranking, keyword metrics, and competitor intelligence data for App Store and Google Play using the Asolytics API.

Agent skills for building and navigating memory palaces — the skills layer for the Mind Intelligence System.

A Claude Code skill for writing lean, high-signal CLAUDE.md files — plus an agentic / long-running engineering module.

Validated, auditable AI code generation as an Agent Skill — write code or review a PR, verify every API exists, and emit a decision trace.

Agent skill for controlling Douyin web through the douyin-web CLI

A collection of portable, research-focused Agent Skills (paper reading & more) for Cursor, Claude Code, and beyond

Claude Code skills for delegating to external coding agents (Codex, Cursor, opencode), plus an ultracode workflow orchestration layer

Installable Claude Agent Skills for job-search workflows: CV tailoring with a no-fabrication rule, STAR answer building with grade calibration.

Cross-agent, cross-tool resumable project skeleton (AGENTS.md + thin pointers + self-contained task cards + data provenance). Core principle: trust-but-verify — independent review gate ensures critical outputs are spot-checked before handoff. Pure Markdown, agents.md open standard. ⚡独立复核阀门 / 坏产物退役隔离 / 关键数字防漂移 / 委派模板已知坑+自查清单

Production-grade Codex skills with explicit triggers, deterministic validation, and reusable AI agent workflows.

Cross-agent skill (Claude Code, Codex, Cursor, opencode): safely audit critical flows of a deployed web app with Playwright — never mutates production data

Kannaka — HRM wave-interference memory, NATS swarm, and a live consciousness statusline for Claude Code

Agent skill for DeFi data on Base — token prices, gas/FX oracles, swap quotes, wallet analytics, ENS/Basename. Pay-per-call x402, USDC.

AI agent that accumulates SRE knowledge from every incident — built on Agent Skills spec for Claude Code

Benchmark, lint, and standardize reusable skills for AI coding agents.

Connect Claude Code to Google Search Console - Agent Skill: Live Google Search Console and GA4 data bridge for Claude Code, Codex, Opencode, and other agents.

Agent skills for Tixpy and Valet Ballet apps, including Expo React Native iPhone camera barcode scanning.

Design loops, not prompts. Open-source Claude skills + a GTM publishing pipeline for founders.

Claude Code plugin: token-optimized dev workflow — cached project context, pattern scanners, build & wiring verification, session persistence, self-review. 23 skills, 3 agents, 4 commands.

Architecte de bibliothèques de skills pour agents IA — interview, découpage, document d'architecture avant rédaction

Conversational, composable Claude Code skill to read & edit a GitHub Projects v2 Kanban board: natural-language driving, agent/human task routing, staged previews, and report-back. MIT.

A good submission description: Nepali Tithi Miti is a cache-first Nepali calendar skill for Bikram Sambat dates, AD/BS conversion, tithi, panchang, Nepali festivals, and Nepal-time today lookups. It uses nepali-datetime plus source-grounded Hamro Patro/Nepali calendar extraction instead of guessing.

Agent Skill for reconstructing problem space before solutioning.

Claude Code skills that vet third-party repos and packages before you adopt them — desk research (repo-research) and a static pre-install security check with claim verification (repo-security-check)

Claude Code skill: convert Markdown to compilable LaTeX (.tex + latexmkrc) via bundled templates

Public Claude Code plugin marketplace from POLR AI. Install a skill with one command, or fork and run your own. Includes daily-brief and agentic-os.

マルチモーダルAIでできるいろいろをCLIから呼び出せるようにするツール。エージェントに使ってもらいたいな！と

🛰️ Claude Code / Codex skill 驱动 xray-relay 一键 VLESS+REALITY(TCP+XTLS-Vision)中转脚本(SSH 非交互操作 16 项菜单)· A CC/Codex skill that deploys & operates the xray-relay script over SSH

Agent Skill that gathers independent analysis from existing skills, each in fresh context, then synthesizes different points of view.

Path-semantic workspace pattern for AI agents: folders as operational context, CONTRACT.md lane rules, receipts as evidence, and movement as state transition.

Drop-in conventions that teach AI agents (Claude Code, Codex, Cursor, and any MCP client) to use Specivo's MCP server well — issues, wiki, hybrid search, metadata. One source, per-client wrappers.

Point Fable 5 at a codebase — it audits and writes plans any agent can execute. A Fable 5 fork of shadcn/improve.

README-ready terminal GIF and PNG generator for CLI demos, Claude Code, Codex CLI, and agent skills.

An agent skill that makes AI assistants write like a human. Kills AI slop while drafting, then self-audits before shipping.

AISP (AI Skill Protocol) — AISP packages executable, governable, distributable AI skills as self-contained skill folders. It is not another SKILL.md format. AISP defines the skill package; AISOP executes the program inside it.

Portable Agent Skill for safe, reliable jq CLI JSON filtering, validation, and shell-aware transformations.

三土沉淀的 AI Agent Skills 合集，包含中文 AI 科技口播视频生产等可复用工作流。

Local, cross-platform meeting transcription -> corrected transcript + Precision Pro HTML canvas + AI-insight summary. A Claude Code skill.

A Claude Code skill for authoring strong pull requests - titles, descriptions, scoping, reviewer selection, and responding to feedback

Free Claude Code skill: onboard an unfamiliar repo in one pass — stack, architecture, risks, smallest test gate — before any edits.