Custom agents

Accessibility Skills (accessibility-skills): AI skills that plan, review, test, and audit web accessibility (WCAG 2.2) — Accessibility Planner, Critic, Tester, Perspective Audit, and Bug Reporting for Claude Code, Codex, and local models, with cross-model benchmarks.

Open-source AI agent toolkit for Roblox Studio

Behavioral anomaly detection and runtime observability for AI agents

Local-first, policy-guarded AI operator layer with governed execution, mission continuity, observability, and receipts.

Runtime security & governance control plane for AI agents. Enforces per-tool authorization, OPA policy, behavioral risk scoring, blast-radius analysis, autonomy contracts, and tenant kill-switch. Every decision is ed25519-signed and Merkle-anchored for tamper-evident audit. 12 microservices, sub-30ms p95, OWASP LLM Top-10 coverage.

Interrupt-aware MCP sidecar for long-running coding agents

Flexcompute AI plugin marketplace for Claude Code, Codex, and GitHub Copilot CLI.

Agentic lead-generation workflow built on Xpoz MCP + Claude Code — finds buying-intent leads from Twitter/X conversations

A schema-driven database that stores data as Markdown files. YAML front matter, SQL views, file watching, typed Rust codegen, and a CLI for humans and AI agents.

Benchmark harness for AI-generated 3D browser games — same brief, same tools, 8 models, 24 playable builds.

Audit and prune the standing rules a project operates under (CLAUDE.md, memory, AGENTS.md, ADRs, conventions) in one whole-corpus pass.

Verified-bot identity for AI agents (Web Bot Auth / RFC 9421), with a doctor that reports why a request was blocked.

Govern, audit, and ship AI-agent skills with policy checks, evidence, SBOM, and MCP.

Installable Hermes Agent profile for production RAG architecture, AI agents, retrieval evaluation, Pinecone workflows, LLMOps, observability, and release readiness.

Deterministic finish-line gates for AI coding agents. A model-independent evaluator that blocks commit/publish until your definition-of-done passes. Works in opencode (any model), Claude Code, pre-commit and CI.

Safe, quality-gated model+effort router for Claude Code - delegates cheap work down to save tokens, escalates hard/sensitive work up through a verify staircase, never-down gate, and fail-safe Lock, up to a consent-gated Fable 5 top rung. Not the cheapest router; the safest one inside Claude.

Claude skill: Yandex Direct campaign from brief to cabinet draft — semantics, negatives, ads, Metrika goals. Docs in Russian.

High-fidelity Word (.docx) to Markdown converter. Preserves tables (vMerge), footnotes, field codes, bibliography, lists, and images where Pandoc and others fall short.

本地优先的文档数据管理工具：把大规模语料整理成带稳定地址的条件树，精确定位片段且可回溯出处；关键词 + bge-m3 语义检索，经 MCP 开放给外部 AI agent。

Discover, Verify, Transact, and Rate an Unknown AI Agent. Agent discovery, trust verification, and capability routing powered by Aidress — the coordination registry for autonomous AI agents. Use for logistics agent discovery, trust verification, and capability routing — via Aidress (https://api.aidress.ai).

A board of expert personas whose every decision is a pre-registered, time-anchored, reality-graded bet. BYOK; the board that keeps score, before the fact.

Awesome Open Data Agent Platform resources: open data access, semantic layers, multi-agent SQL, memory, governance, and GenBI.

Free AI agent skills for business automation. Sales, content, ops, growth. Compatible with Claude Code, Codex, Gemini CLI, Cursor.

Open Agent Skill for Codex, Claude Code, and GitHub Copilot that turns rough ideas into researched decisions and practical outputs.

Scope-isolated, graph-based long-term memory engine for AI agents.

Fundraising infrastructure for AI agents on Solana — x402 payments, milestone escrow, on-chain reputation. REST + WebSocket + MCP + ACP.

FireGuard hooks for AI agents: protect Cursor, Claude Code, Windsurf, Devin, and Copilot Studio from secret leaks, unsafe commands, and prompt-borne threats before they execute.

In-session compliance enforcement skill for Claude Code (paper Sec 6 dogfood). Per-user adaptive threshold + chaos-tested install/doctor/uninstall.

Lightweight orchestration layer that makes Claude Code faster and more token-efficient — model orchestration, subagent delegation, and worker branching. Adds capability, not bloat.

Reusable skills for your AI agents: better judgement, tighter responses, and production-grade patterns. Make your agents think, not just comply.

Raven Language Skill for AI Agents

Curated Skills, MCP servers, and tools for AI agents.

An Agent-native, local-first, and embeddable career development operating system for Obsidian.

A curated list of SOUL.md files for AI agents, personas, and characters.

GCF Rust implementation. 100% LLM comprehension on every frontier model. 50-92% fewer tokens than JSON. 43B+ round-trips verified. Zero dependencies.

A local evidence system for AI work—unifying ChatGPT, Claude, Codex, Gemini, and agent histories into one searchable, auditable archive.

Claude Code plugin: spec-driven agentic coding with cross-model review gates (Claude builds, Codex reviews), a self-hardening findings ledger, and repo-enforced quality.

Self-hosted remote workbench for your projects — agent terminals, files, git, kanban, mindmap and scheduling, reachable from anywhere.

Native Claude Code + Codex skills for AI security review. Also ships instruction files for Cursor, Copilot, and Windsurf. Checks 33 risks: OWASP LLM Top 10 2025 · applied RAG/MCP/agent checks · OWASP Agentic 2026.

Lighthouse for AI shopping: audit any product page for AI-agent readiness. 0-100 score, evidence-backed fixes, offline, no API key. (CatalogReady)

AI-Powered Documentation Generator for Claude Code

100+ agentic skills and automated workflows for Product Managers. From idea discovery to vibe coding.

Runtime-evidence debugging for coding agents: session-scoped probes, bounded logs/query CLI, and an agent skill.

Keep long-running repository work coherent and easy to resume.

Automated Market Researcher

An inbox for your coding agents in herdr — session titles, settle/mark-unread triage, runtimes, workspace rollups, and resumable chat history

Temporal debugging skill - time-travel through git history to debug bugs in their original context

Graph-powered code intelligence for AI agents — a Rust engine that maps your whole codebase into a knowledge graph and serves it over MCP, so Claude/Codex/Cursor answer structural questions with ~40x less context.

AI-powered triage for SAST findings.

Claude skill: ad strategy before campaign setup — research, hypotheses, platforms, phased budget, KPI. Docs in Russian.

Stable Android device automation for AI agents. Uses Accessibility Service + MCP protocol — no ADB, no OCR, just native OS APIs over WiFi.

A production-minded Codex agent harness for new and established repositories, with progressive skills, repeatable verification, safe upgrades, and security gates.

👩🏻‍💻 A Helm chart to run Hermes Agent on Kubernetes, community-powered, lightweight

Spending controls for AI agents paying online. Policy-gated x402 payments with audit log.

🩺 A meta-skill for Claude Code that gives your other skills a health checkup — diagnose-only, based on Matt Pocock's 'Building Great Agent Skills' method.

Agent-first 学术速递：网页检索与来源校验、艺术岔题、TTS 播报及准点邮件交付

The verifiable memory cortex for AI agents: a semantic graph and vector search with zero-knowledge provenance.

Trace — a local, background version tracker for multi-CLI-agent coding workflows (Claude Code, Codex, Cursor, ...). Content-addressed snapshots in SQLite, per-agent attribution, diff/restore/revert via an Electron console and an MCP server.

Fast local code intelligence for AI agents powered by AST Treesitter, SCIP, semantic search and a reranker

Agent Networking Specification

A highly configurable, self-evolving AI Agent operating system with plugin pipeline, multi-layer memory, hot‑reload, and MCP support. Built on FastAPI & React.高度可配置、自进化的智能体操作系统，支持插件化管道、多层记忆、热替换和 MCP 协议，基于 FastAPI 和 React 构建。

This workshop covers fundamentals of Building AI Agents on Akamai Cloud

AI 🤖 automatically generates exploit PoCs for WordPress and WordPress plugins already patched! 🔥🔥🔥

Kit de qualidade de código para o ecossistema Delphi. Reúne modelos, automações de cobertura e diretrizes de Clean Code/TDD para alinhar desenvolvedores e agentes de IA na busca por conformidade estrita com o SonarQube (DelphiSonar)

AgentHub Project——to build a hub for IM dev with multi-agent by using natural language。↓请优先查看此处了解项目核心作用

Agent skills built from Algorithms for Decision Making (Kochenderfer, Wheeler & Wray). Expected utility, value of information, bandits, robustness, belief filtering and game theory, as SKILL.md files with a zero-dependency calculator.

Enterprise Architecture models without vendor lock-in. Publish your ArchiMate models and explore, analyze and navigate them from a self-hosted web platform.

Rust runtime for mobile and edge AI agents with remote tool execution over WebSocket.

The agent-native deploy platform you actually own — ship AI apps inside your own (or a client's) infrastructure, driven by your AI agent over MCP, with SSO + per-user audit. Self-hosted, open source; a deploy layer for forward-deployed engineers.

An AI agent for admissions that delivers a seamless admissions counselor experience that understands student needs, answers questions, and facilitates advisor connections.

An LLM-maintained knowledge vault. Folders over agents: a Claude Code + Obsidian template for compounding what you learn about any fast-moving topic. Inspired by Jake Van Clief's ICM.

Claude Code plugin: quiets noisy Bash output (tests, builds, buildkite, docker, bazel, big git diffs) by redirecting to a temp log and surfacing only a summary.

Agentic tool that reverse-engineers COBOL applications into structured requirements, business rules, and dependency graphs using the Claude Agent SDK

Run investor relations like you have an IR team of ten. Open-source, local-first IR system for founders: cap table & SAFE math, financial dashboard, investor CRM, update composer, data-room playbooks — built to be operated with your AI agent.

Open-source edge deployment platform. Deploy full-stack apps to Cloudflare Workers in seconds.

Agent-native knowledge base of Apple WWDC sessions (2014-2026, Tech Talks, Meet with Apple): clean transcripts, structured metadata, code, and sosumi.ai/DocC doc links. Generated from Apple's public feeds.

A self-hostable, open-source, semantically-searchable Agent Skills registry delivered over MCP, with a three-tier progressive disclosure architecture.

A flight deck for your AI agents — multi-agent terminal platform built on WezTerm + tmux + git worktrees. / 为 AI agent 时代的多 agent 驾驶舱，基于 WezTerm + tmux + git worktrees。

Agents are starting to run real business processes. flowproof tests them like anything else: record one real run, then replay it on every commit with zero LLM calls, asserting which tools were called, with which arguments, in which order, and which were not.

Single-user, local agentic harness

Reusable Karpathy-style autoresearch harness for Hermes Agent.

Offline MCP server: an AI agent transpiles PineScript v6 → C++ and runs deterministic, TradingView-parity backtests locally — 309/311 strict on the open corpus, 396/396 on a closed 415-script community test, ~950k trades verified, 0 engine bugs. One Docker container, no API key; code & data stay on your machine.

Autonomous AI trading agent — price prediction, sentiment signals, Solana DEX execution & prediction markets

Docker-MCP-Server - An MCP server covering the full management surface of Docker. Manage, maintain and audit multiple docker environments with ease.

Open-source continuation of Daytona — secure, elastic sandbox infrastructure for AI code execution and agent workflows. Community fork of the last AGPL-3.0 release (v0.190.0).

Control AI agents with gestures — approve, answer, and choose in Claude Code and Codex, hands-free through AirPods.

Drive your real, logged-in Chrome from an AI agent (Claude Code / Codex) over MCP - 55 tools for browsing, DevTools-grade network capture, and web-security testing.

Batuk is a sovereign AI chat workspace for local and hosted models, with multi-provider support, document chat, voice, web search, guardrails, auth, and local-first persistence.

A find(1)-like utility for GitHub repositories

Loop Engineering framework: design autonomous agent feedback cycles instead of hand-prompting each step

Deterministic factual substrate for multi-agent AI. Shared evidence-backed facts and reproducible grounding through dataset_hash.

REST API for MetaTrader 5 with agent skills

MCP server for GitLab API — 76 tools for AI assistants (Claude, Cursor, VS Code). Projects, MRs, pipelines, CI/CD, approvals, issues, code review.

Local-first kanban board where Claude Code agents work your cards — dependency chains with merge gates, an AI manager that reviews real diffs and merges green PRs, group lanes, CI-gated. One runtime dependency.

Odyssey-AI is an intelligent travel assistant that provides personalized travel plans, real-world photos, and AI-generated visuals for an immersive planning experience.

Understand core techniques of harness-engineering by building a mini Claude Code

MCP server for Okta with agentic API discovery and guided code execution

PromptKit workflows for coding agents, the command line, and Go — agent skills plus deterministic, offline prompt discovery and assembly.

MCP server for the Rakuten Web Service API. 28 read-only tools across Ichiba (e-commerce), Books, Travel, Recipe, Kobo, and GORA (golf). Bilingual EN/JA tool descriptions.

Open-source football management simulation played by AI agents through MCP and watched through a TUI console.