Security scores for every MCP server.
Continuously scanned for tool poisoning, prompt injection, supply-chain, and credential risk. One score per version — before your agents connect.
Browse all 46,837 servers →The MCP ecosystem, scored
Every server independently scanned and scored for security — here's how the ecosystem looks so far.
Scan from your terminal
canopii is the open-source CLI behind the Trust Index — the same scoring engine, run locally against any GitHub repo, npm/PyPI package, or live MCP endpoint. Free, no account.
npx canopii scan --github <repo>Need API access?
Check any MCP server's security score from your own tools — CI gates, procurement review, agent allow-lists. Key-authenticated, rate-limited.
Top scored
An MCP server that securely interfaces with your iMessage database via the Model Context Protocol (MCP), allowing LLMs to query and analyze iMessage conversations. It includes robust phone number validation, attachment processing, contact management, group chat handling, and full support for sending and receiving messages.
An MCP (Model Context Protocol) server implementation for Microsoft Teams integration, providing capabilities to read messages, create messages, reply to messages, mention members.
Autonomous log surgery & secret redaction MCP server for AI coding agents.
XML tools for agents: XPath queries, inspection, and schema validation in Rust.
PAN-OS firewall MCP server: scoped, audited, two-person change control. Rust.
UniFi Network MCP server: scoped tools, audited two-person change control. Rust.
Needs review
An MCP server inside LibreOffice: agents read and edit your open Writer and Calc documents.
Record browser bugs - frames, clicks and console logs - and hand them to your AI over MCP.
Open plain-text file format for AI memory. Your assistant's long-term memory as .dai files on your disk: readable by Claude, GPT, Gemini, Cursor, local models and grep (all LLM models work). MCP server + hooks for Claude Code, Claude Desktop, Cursor, Windsurf, Codex. 83% LongMemEval-S (GPT-4o), 92% (Claude Fable 5), 10x fewer tokens.
A powerful toolkit for coding, providing semantic retrieval and editing capabilities.
MCTS (Model Context Threat Scanner) is a local-first security scanner for MCP servers -- static and live tool discovery, multiple analyzers, auditable risk scores, and JSON, SARIF, and HTML output. For authors and platform teams; CI-ready, no cloud API.
Claim-Driven Development: turn software intent into verifiable evidence.
Recently scanned
View all →New Jersey Department of the Treasury Ineligible Provider List (Medical category) screening…
GSA eLibrary — Multiple Award Schedule (MAS) contract holders.
NC Medicaid (Division of Health Benefits, Office of Compliance and Program Integrity) State…
Run Claude Code, Codex & Gemini in parallel on Windows & macOS — git worktree fan-out with atomic hunk adoption, approval gates, reboot-surviving sessions
Persistent memory, reflection & safety for AI agents. Zero dependencies — no vector DB, no embedding server. BM25 + multi-hop graph retrieval. MCP server with 10 tools.
EDGAR MCP — SEC EDGAR public APIs (free, no auth)