#mcp

Official @agent-nexus/cli + @agent-nexus/sdk — public mirror of packages in NexusGPT/nexus. Source of truth lives in the monorepo; this repo is auto-synced and used to publish to npm with provenance attestations.

Local inventory of the AI agent surface on your machine. No account, no network, no upload.

One MCP server for all web data — search, scrape, crawl, proxy, and AI research in a single npx install. Works with Claude, Cursor, and any MCP client.

One config. Every AI agent. Zero drift. Idempotent setup for OpenCode, Claude Code, Cursor, Windsurf, and Copilot.

A drip-feed verification pipeline as an MCP server: phases, steps, and a cursor the agent cannot move.

A framework for giving any AI agent persistent, browser-trainable SSM memory — Mamba models, swappable LLM bridges, and online distillation. The hippocampus behind BuilderForce.ai.

A discovery layer for everything in the Stellar Ecosystem

Claude Code plugin that delegates code review, adversarial design critique, and isolated implementation to Kimi Code over the Agent Client Protocol (ACP). Security-first: Kimi works in an isolated workspace and hands back a Git patch.

MCP Server for YouTube API, enabling video management, Shorts creation, and advanced analytics

No repository description published.

MCP server + CLI that puts a real human in your coding agent's loop. It publishes work mid-task, a reviewer pins what's wrong and returns a verdict, and the agent iterates until approved.

Bootstrap library for Go MCP servers — one Run() call instead of ~80 lines of boilerplate

Standalone DAG workflow engine for Go — 15 step types, MCP integration, durable Postgres execution (SKIP LOCKED), templates, approvals, crash recovery

core2ai provides the shared MCP runtime and code generation infrastructure used by api2ai and db2ai to build AI tools from APIs and SQL queries.

Identity-scoped schema selection for text-to-SQL: give the model only the tables this caller may read, and 75% fewer prompt tokens.

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

Stateful, structured, safe command execution for AI agents - over local shells, SSH, and Docker.

MCP server with 30 tools for block chain data and ops— setup, token lookups, wallet balances, live chain queries using micro payment on x402.

用最少的代码复刻 OpenClaw 的 agent 运行原理

The identity provider for AI agents — registry, scoped delegation, in-path MCP enforcement, instant revocation, tamper-evident audit.

AI Agent Governance Runtime - Deterministic control plane for AI agents.

A thin MCP client for humans and agents — run any Model Context Protocol server, call its tools, read its stderr, and reconnect after each edit. REPL, headless CLI, or an MCP server your agent drives. No client config edits, no restarts.

dst (data serve tool) — open-source semantic layer and MCP server between the AI your team uses and your data warehouse. Governed answers with the SQL, a grade and a receipt.

AI agent that lives in your chat apps. Controls your machine, remembers your conversations, searches your documents — with guardrails so it doesn't go rogue. OpenClaw without the bloat

Self-improving context & memory for Claude Code and OpenAI Codex. One canonical knowledge store injected only when relevant — hooks push budgeted context, an MCP server serves depth on demand, and it learns lessons from your sessions. Agent memory + context engineering as a single static Go binary.

No repository description published.

A fast, scriptable CLI for the full Chatwoot API — 144/144 operations, profiles, keyring auth, MCP server, agent guardrails

Interactive question boards & HTML idea visualization for AI coding agents (Claude Code, Codex, …) — ask users structured questions in the browser, await submit, get JSON. Zero deps.

MCP interface to BioPython

Static analyzer that turns a Python package's source into a queryable code graph (API surface, dependencies/architecture, impact, change-review, RAG).