@AgentSafe-AI
基本情報
Scan MCP servers for security risks before your AI agent trusts them. Detects prompt injection, supply chain attacks (including the LiteLLM 1.82.7/1.82.8 backdoor), excessive permissions, arbitrary code execution, typosquatting, and tool shadowing. Add to your .mcp.json and let y
以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。
{
"mcpServers": {
"tooltrust": {
"command": "npx",
"args": [
"-y",
"tooltrust-mcp"
]
}
}
}Returns the full catalog of security rules used by the ToolTrust scanner, including rule IDs, titles, and descriptions. Useful for understanding what the scanner checks for.
Look up historical security risk grades for an MCP server from the public ToolTrust Directory. Accepts the kebab-case name of the server and returns its full JSON scan report, or 404 if not found.
Reads the user's Claude Code MCP configuration and scans all configured servers in parallel. Searches for .mcp.json in the current directory, then ~/.claude.json as fallback. Returns a summary report with scan results for each server. Servers that fail to start are reported with an error note; scanning continues for remaining servers.
Connects to a live MCP server via standard input/output (stdio), parses its tools, and scans them for prompt injection, data exfiltration, and privilege escalation risks. Returns a risk report with gateway policies (ALLOW, REQUIRE_APPROVAL, or BLOCK) for each tool.
Scan a list of AI agent tool definitions for security risks. Accepts an MCP tools/list JSON payload and returns a risk report with gateway policies (ALLOW, REQUIRE_APPROVAL, or BLOCK) for each tool.
Tooltrust Scanner audits MCP servers for prompt injection, data exfiltration, and privilege escalation before an AI agent trusts them. It is designed for developers who integrate MCP servers with AI agents and need a security preflight check.
Add Tooltrust as an MCP server in your configuration using the command npx -y tooltrust-mcp. Then ask your agent to run the MCP tools tooltrust_scan_config to scan all configured servers or tooltrust_scan_server to scan a specific server. A full catalog of MCP tools is documented in the usage guide.
exec, network, db, fs)It catches prompt injection, excessive permissions, supply‑chain vulnerabilities, privilege escalation, typosquatting, tool shadowing, insecure secret handling, and missing security configuration on risky tools.
Add it to your MCP client configuration using `npx -y tooltrust-mcp`, then ask your agent to call `tooltrust_scan_config` or `tooltrust_scan_server`.
Tooltrust Scanner runs via `npx` (Node.js), so a Node.js runtime must be available on the machine where the MCP client runs.
Yes. A live UI is available at tooltrust.dev where you can browse the public registry, review findings in the browser, and compare grades across servers.
Yes. It includes an urgent security update that detects and blocks known compromised MCP‑related package versions, including the LiteLLM / TeamPCP exploit.
Remote MCP server for Mythex, an AI app builder. Claude, Cursor or Codex get a cloud sandbox to build full-stack apps: files and bash, live preview, Postgres database, file storage, secrets, checkpoints, browser tests an
Pugi is an AI product team that builds, checks and publishes websites and apps. Its MCP server gives agents a passive website security check graded A+ to F (headers, TLS, cookies, exposed keys in scripts), free with a no
DrUrls audits websites for SEO, and its MCP server lets Claude, ChatGPT, Claude Code, Cursor or VS Code work with those audits directly: list your sites with their latest score and grade, read each issue across SEO, perf
Routebase keeps your API's design, docs, contract tests, mocks and monitoring in one living OpenAPI spec so they can't drift apart. Agents work on that same source of truth, under your team's permissions.
Know when your AI agents, cron jobs and CI pipelines fail, stall or go quiet. Your agent sets up monitors from a plain description, tells you what is broken and why, reads run history and traces, and writes notes on inci
コメント