What problem does Carapace solve?
Carapace addresses two critical risks of autonomous AI agents: prompt injection (tricking the agent into executing unauthorized actions) and goal drift (the agent losing focus over time).
@jdesl55
Armor for your AI agent. Carapace is a local-first security system that protects autonomous AI agents from prompt injection and goal drift. It verifies actions against user-defined rules before execution, uses rotating HMAC keys that injected prompts can't forge, re-reads goals e
使用下面的配置,将此服务器添加到你的 MCP 客户端。
{
"mcpServers": {
"carapace": {
"command": "carapace-mcp",
"args": []
}
}
}Security checkpoint. Validates actions against your rules, returns pass/block verdict with rotating key.
Goal journal. Returns your goals, priorities, and constraints. Detects drift by comparing agent activity against configured categories.
Action logger. Records what the agent did, flags unverified sensitive actions.
Security posture summary. Health status, action counts, key rotation timing, drift level.
Session grader. Analyzes all actions against goals, produces a scorecard with grades, highlights, and improvement insights.
Carapace is a local-first security system for autonomous AI agents. It plugs into any MCP-compatible agent (OpenClaw, Claude Code, Codex, and others) and provides security verification, goal anchoring, drift detection, action logging, and session grading — all running on your machine with zero cloud dependencies.
Install via Git and npm, then build the project. Add the server to your agent's MCP configuration (e.g., ~/.config/openclaw/mcp.json) by specifying the path to the built index.js. On first run, Carapace creates ~/.carapace/ with configuration, database, and security secret. Edit ~/.carapace/config.json by hand or use the optional Carapace Dashboard.
Carapace addresses two critical risks of autonomous AI agents: prompt injection (tricking the agent into executing unauthorized actions) and goal drift (the agent losing focus over time).
Every 30 minutes, Carapace generates a cryptographic verification key using a secret that never enters the agent's context window. The real agent has the key; a prompt injection cannot forge it.
Carapace requires Node.js and runs on any platform where Node.js is available. It uses TypeScript, the MCP SDK, better-sqlite3 for local logging, and HMAC-SHA256 for key verification. It has zero external dependencies for core security operations.
All data lives in `~/.carapace/` on the local machine. This includes `config.json` (security rules, goals), `logs.db` (action log), `.secret` (HMAC signing key with 600 permissions), and `insights.md` (session learnings).
Carapace communicates over stdio via the Model Context Protocol. The HMAC secret is created with owner read/write permissions only and never enters the agent's context window or is transmitted over any network.
🚀 The fast, Pythonic way to build MCP servers and clients.

Yuppsy MCP: build and publish real estate websites from your AI assistant. Yuppsy lets your AI assistant create live websites for real estate agents and individual property listings. You don't need a designer or any code

Couples therapist Figs O'Sullivan (LMFT, 17 years, endorsed by Sue Johnson, creator of EFT) inside your AI. Ask relationship questions answered from his clinical writing, name the pattern in a fight, get a three-step rep

Health by Dr Brad works out a health plan from your own results: which screening is due at your age, and evidence-based suggestions, each with the reason and the published guideline or study behind it. Your record is one

Request dedicated senior developers to work on your project, write job descriptions, and prep technical interviews — all without leaving your Claude chat. Just describe what you need, and Lemon.io MCP does the rest. 1–3
评论