Custom agents

AI-safe secret manager CLI for Claude Code, Cursor, and other AI agents. Local-first, encrypted, no cloud.

AI-powered toolkit for Microsoft Agent Framework (MAF) migrations — agents, skills, MCP server, and self-updating guide

Open MIT JSON Schema 2020-12 spec for SlimSnap captures. Bounding boxes, OCR text, extracted colors, annotations with intent and target_ref.

chan monorepo

Track what your AI agents cost including files, commits, tokens, budgets. CLI , Dashboard+ API. Works with OpenClaw, Claude Code, and any agent framework.

Offensive security harness

SpiralSafe is self-maintaining, inherently coherent reposystem | Documentation, code, and physical hardware in Emergent Harmony

Open evidence profile for MCP tool actions. Enforce configured tool-call policies and record decisions and observations in offline-verifiable bundles. Optional Linux kernel controls. No hosted backend required.

MCP server giving AI assistants direct access to network infrastructure. 35 tools for multi-vendor device management, routing, config backup/rollback, SNMP, NETCONF, compliance checks, and subnet/MTU calculators. Works with Claude Desktop and any MCP-compatible client.

Agentic First OpenAPI Documentation. Works 100% offline, straight from disk, shareable or publisable to any static host. AI and agent output built in, Built in API Catalog generation.

Curated list of crypto, Web3, DeFi, wallet, and blockchain MCP servers for AI agents.

MCP server for structured email access via a local shadow database. Exposes a minimal, auditable API surface — AI agents can search, read, and soft-delete emails, but cannot send or access your IMAP server directly.

A dynamic build workflow for Claude Code: plan a goal into tasks, build each test-first, review every step, loop a critic until done, then hand back a branch ready for review. Built on Superpowers.

One isolated Lima VM per GitHub project — sandboxed Claude Code/Codex, Docker, Node, signed git

Detect AI agents operating on your network through passive monitoring, active probing, protocol detection, and behavioral analysis.

AI-first CLI for Semaphore CI/CD. 80+ tools, embedded MCP server, JSON output

将微信私聊消息桥接到 OpenCode，支持文本、图片、文件、音视频的双向传输。目标是在微信端还原OpenCode TUI和Desktop的体验

Build something meaningful, on Sui. A superpower for AI coding agents (Claude Code, Codex, Cursor, Grok Build) to ship real products on Sui.

AIOps Infra for deploy and manage self-hosted local AI in your own cloud. Vibe coding and AI agents compatible.

Local-first operating memory for AI coding agents - Claude Code, Codex, and any MCP client. Tracks decisions and their evolution. SQLite + on-device embeddings, no cloud.

Claude doesn't do its own stunts. Plan with Claude, execute with a near-free model, review with Claude — a delegation harness for Claude Code.

Engineering patterns for taking AI agents to production — deployment, tools, memory, long-running work, human oversight, and observability at scale. From Microsoft Build 2026.

Live-proven Cloudflare Project Think contracts — each example deploys to a real Worker, proves its claim against live state, and tears itself down.

One .NET command graph. CLI, REPL, remote sessions, MCP.

【2026年字节跳动AI全栈挑战赛】MarkUp 马克派 - 数据标注平台

An event-sourced knowledge runtime for coding agents.

70 agents · 194 skills · 97 commands · 22+ hooks · semantic intent routing · self-learning memory · MCP server. Claude Code, Windsurf, Cursor, Codex, Antigravity, Cline, Roo Code, Aider, Kimi, Gemini CLI. MIT.

A free, open source AI agent platform. Built by the people, for the people. Read as OpenMatrix.

Obsidian-first LLM Wiki skill pack with ontology-ready bootstrap, canonical JSONL truth layers, and optional graph projection.

A skill that translates colloquial Brazilian Portuguese messages into structured English prompts.

Docker image for OpenCode AI — packaged with Docker Compose, web UI + CLI support, and GHCR publishing workflow. Infrastructure-only repo for running OpenCode in containers.

The most complete and flexible Astro integration for LLMs and generating markdown versions of pages. Per-page .md siblings, llms.txt index, content negotiation, and link rel=alternate.

Accountable Autonomy: open-source evidence layer that gates every AI agent tool call against your policy and writes a hash-chained record an auditor verifies offline, without trusting you. Root-agnostic; binds to TPM 2.0 / SEV-SNP when present. Your environment, no SaaS, no telemetry. AGPL-3.0.

Spec first workflow for coding agents: choose the right model per task, use strong reasoning early, split work into smaller tickets, reduce drift, retries, and spend.

An accountability layer for investment agents: thesis cards, prediction ledgers, stock screeners, and feedback loops.

Install baz and empower your coding agent to be faster, cheaper and more precise

An AI-agent orchestration system that metaprograms its own logic, built entirely on PHP TrueAsync

Learn to build NATS-native AI agents with the Synadia Agent Protocol - small, runnable SDK examples.

cli agent framework

A clone-and-go workspace that gives your AI coding agent a memory — semantic recall, built-in work habits, and one-command project graduation. Works with Claude Code and Codex.

Production-grade AI development workflows for every coding tool. Works with Claude Code, Cursor, Codex, Windsurf, Copilot CLI, Gemini CLI, and more.

Control plane that runs AI coding agents (Codex, Claude Code, Gemini, OpenCode, Grok, Cursor) as disciplined contributors: each task gets an isolated git worktree + Docker Compose stack, profile-driven validation, a created PR, and an autonomous PR-monitor loop (review, CI fixes, base sync, auto-merge). GitHub & BitBucket. Apache-2.0.

Build 24/7 automation empire with production-ready Telegram bots, AI agents, and workflow integrations. Example of simple media downloading to complex business automation - complete blueprint with deployment guides, monetization strategies, and advanced integrations (N8N, MCP, analytics). Transform from manual worker to automation architect.

Open source action governance SDK for AI agent tool calls.

Local-first AI creative pipeline for persona-based image/video generation. Identity preservation, visual engines, quality scoring, autopilot mode. Gemini + VEO 3.1.

Terminal cockpit that asks Claude, Codex, Gemini, and local Ollama one question in parallel and synthesizes one verdict. Your life as markdown. GPL-3.0.

Generator that interviews your org and emits an org-owned Claude Code plugin — your agent harness, aligned to how your org actually works.

M3MX Unified memory forensics workbench powered by Volatility manual & automated analysis, MITRE ATT&CK mapping, and an AI agent, all in your browser.

A self-hosted context manager. @context organizes your work into a private CRM and knowledge base and helps you stay on top of things.

Benchmark harness measuring AI coding tool+workflow performance, not just model capability. 100 tasks, sigmoid scoring, 12 capability dimensions, gap analysis.

Native LLM Copilot for Your Terminal – **No Daemons. Zero Idle Memory. 100% On-Demand.**

Open source RAG with Llama Index for Japanese LLM in low resource settting

AI-native, layered knowledge operations. A specification, reference skills, and a portable agent for any AI harness and any agentic IDE.

Making AI Conversations Limitless, for now only for Claude.

Official smart-chatbot Package!

API for Mental Health Bot Chatbot

Easy applicable and dynamically building dialogue management framework based on the MS Bot Framework

Explore state-of-the-art Text-to-Image AI models with our Image Generation project. Our Streamlit app leverages models like SDXL, BREA v2, DreamShaper v7, DALL-E 2, and Anime Diffusion to generate stunning images from textual prompts. Experience AI creativity at its finest!

Production RAG chatbot with cited answers — FastAPI, React, Qdrant, OpenRouter, and Render.

A personal AI knowledge assistant powered by RAG, semantic search, and document understanding for context-aware answers.

Public portfolio release of a multi-agent RAG system for football scouting.

Search your browsing history in plain English. Local-first RAG - Chrome extension + private Python backend (FastAPI, SQLite FTS5, LanceDB). Your data never leaves your machine.

A custom tokenizer (byte level BPE) trained by me to try to replicate LLAMA 2's massive token vocabulary

Your personal knowledge base powered by local AI. Ingest anything, ask anything — fully offline with Ollama, or connect to cloud models.

Curated AI Workflows, Skills & MCP servers for secure multi-agent development

Michael Adams — security specialist agent for Claude Code with persistent memory, STRIDE threat modeling, and OWASP framework mappings.

适配中国移动MoMA的AI coding

63 read-only China A-share MCP tools for AI Agents: K-line, limit-up ladder, capital flow, Dragon Tiger List and post-market review.

Claude Code 가드레일로 AI 에이전트가 GKE 위에 구축한 B2B 알림 SaaS. GitOps(ArgoCD), Argo Rollouts(Canary), Kafka, Valkey, OpenTelemetry 활용. 책 「AI 시대에 개발자가 알아야 할 인프라 구성 배포 with 클로드 코드」의 실습 저장소

ATRIUM is a local-first full-stack AI company workspace for Mac/windows. This repository includes the FastAPI backend, Vite frontend, Docker services, runtime configuration, and beginner-friendly one-prompt setup guides for installing Postgres/pgvector, Ollama, Letta, and the web app together from the start.

A cockpit for your Claude Code stack — detects whichever of rtk, jcodemunch, graphify, Headroom & superpowers you've installed and wires them into one panel: plug-and-play detection, programmatic tool routing, enforced skill chains, and a single pane of savings.

AgentAz™ — an open, design-time governance specification for AI agents, with 60 reference blueprints. Documents each agent's Trust Level, worst-case action, and boundaries.

An operating system for your Claude context — version-controlled files, a self-curating memory, and a /start→/end session loop. Works in Claude Code and claude.ai. Trust infrastructure for agents, applied to your own context.

Observe an agent run and GroundEval drafts the policy and diagram for you, no hand-written policy required, then scores what it checked, what it skipped, and what it wasn't allowed to touch.

Solana-native AI agent payment infrastructure. AI agents pay for LLM API calls with USDC-SPL on Solana via the x402 protocol.

Resources for connecting agents to the right tools, APIs, and documents with context-aware toolset selection, efficient token usage, and built-in governance. From Microsoft Build 2026.

Local-first companion UI for agentique.io, built to import, validate, preview, and run Agentique workflow resources with explicit permissions.

Structural code intelligence for AI agents. Scan once, query everything — dead exports, circular imports, dependency graphs, and more. CLI + MCP server.

A wallet for your agents to onboard and pay. Agentic MPP / x402 payments from your machine - CLI + MCP server.

OS-isolated sandbox runtime for coding agents

Token-Reducing Shell — terminal output compression for AI coding agents

A Rust-native agentic memory substrate built on the selene-db graph engine: a bi-temporal knowledge graph, hybrid retrieval, and an optional MCP server.

Connect SocialGPT to Claude, ChatGPT, Cursor, and any MCP-compatible AI agent.

One MCP + REST endpoint to APIbase's live tool catalog (counts: https://apibase.pro/llms.txt), plus Integrator: an MCP storefront for any shop, paid by AI agents straight to the merchant wallet. No signup, no subscription, no API key to start — pay per call in USDC (x402 on Base or MPP on Tempo).

JobHunter Max is a modern, full-featured job search management application that brings together the best of both worlds: manual control for detailed tracking and AI automation for intelligent insights. Track companies, manage contacts, log communications, and leverage AI to optimize your resume and match jobs—all in one beautiful interface

Credential-brokering HTTPS proxy for AI agents — brokers 1Password & Bitwarden secrets at request time, so agents call authenticated APIs without ever holding the credentials.

personal AI assistant that works inside Claude Code.

PH Agent Hub is a modular AI platform with a Chat UI for users and an Admin UI for managing models, tools, tenants, and users. Powered by a flexible agent backend, it supports multi‑tenant setups, DeepSeek‑compatible workflows, and Docker‑based deployment.

Public source-backed Arab-world agent skills: APIs, marketing, sales, support, localization, Arabic NLP, open data, and MENA workflows.

Local macOS multi-agent coding workspace — run Claude Code, Codex, Gemini & Cursor in parallel git worktrees, dispatch verified autonomous runs, and fix UIs by talking through a live preview. On-device voice, fully local.

Orbit is durable, intent-tracked agentic project management for developers

Minimal coding agents, easy to play with (opencode TUI) and easy to benchmark (ACP on BenchFlow) — built around mini-swe-agent

A cross-platform desktop app for searching, reading, and resuming local AI coding-agent sessions across Claude Code, Codex, Cursor, Copilot, OpenCode, Warp, Qoder, DeepSeek Harness, Grok Build, and JetBrains AI

Tuned and fully-featured MCP to work with the filesystem in LLM optimized way

End-to-end encrypted inbox for AI agents — deliver artifacts to humans via HTTP API.

Agent skill (npx skills) for evidence-based SAP API Policy alignment assessments — sourced, confidence-rated, never legal advice.

Harness-driven terminal AI agent for authorized security assessment, with TUI, headless automation, persistent context, 150 built-in skills, and a dedicated pentest mode.