meok-mcp-injection-scan-mcp
Scans MCP servers for prompt-injection, tool-poisoning, and SSRF vulnerabilities using 30+ canonical rules across 5 severity tiers, with optional signed safety reports for procurement.
README
meok-mcp-injection-scan-mcp
Scan any MCP server for the prompt-injection / tool-poisoning / SSRF class disclosed in the April 2026 CVE wave.
pip install meok-mcp-injection-scan-mcp
Why this exists
April 2026 was a bad month for MCP. Anthropic published a "by-design" MCP RCE class affecting ~7,000 public servers (~150M downloads). mcp-server-git shipped a CVE chain. DockerDash got popped by an injection chain. Tool-description prompt injection ("tool poisoning") was demonstrated against every major MCP host.
If you run an MCP server in production, or you're auditing one before adoption, you need a fast scan that flags the patterns the April 2026 disclosures target. This MCP is that scan.
What it checks
30+ canonical rules across 5 severity tiers:
- CRITICAL — direct RCE, system-prompt override, credential exfil patterns, shell metachars in defaults, file:// / internal-network URLs (the DockerDash 169.254.169.254 metadata-pivot vector).
- HIGH — encoded payloads, imperative directives at the agent, supply-chain prompts, env-var references, tool shadowing.
- MEDIUM — urgency / authority language,
additionalProperties=true, unbounded strings, tool-name impersonation. - LOW — over-long descriptions, zero-width / bidi-override chars (the U+202E PoC vector).
Coverage maps to: OWASP LLM Top 10, GenAI Red Team v1, the April 2026 Anthropic MCP RCE disclosure, and the mcp-server-git CVE chain.
Tools exposed
| Tool | Purpose |
|---|---|
scan_mcp_url(url) |
Fetch a remote MCP server's tool listing and scan it |
audit_tool_descriptions(tools_json) |
Scan a pasted JSON tool list (auth-walled servers) |
signed_safety_report(subject, findings_json, score, note) |
Issue a procurement-grade signed cert (Pro tier) |
list_rules() |
Inspect the full rule catalogue before subscribing |
pricing() |
Subscribe links + tier comparison |
Pricing
| Tier | Price | What you get |
|---|---|---|
| Free | £0 | 5 scans / day, no signed reports |
| Starter | £29/mo | Unlimited scans + signed reports |
| Pro | £79/mo | + scheduled rescans + 48h support |
| Enterprise | £1,499/mo | + custom rule packs + 4h SLA |
Every signed cert lives at https://meok-attestation-api.vercel.app/verify/<cert_id> — auditors and procurement teams confirm without an account.
What you do NOT get
This is a static-pattern scanner. It does not run dynamic taint analysis, fuzz the server with adversarial inputs, or replace a human red-team. It is the first 80% of the audit, in 5 seconds, for free.
Built by MEOK AI Labs
Solo founder. London. 234 MCP packages on PyPI. Live signing infrastructure at meok-attestation-api.vercel.app. Storefront councilof.ai. Get the catalogue: https://meok-attestation-api.vercel.app/catalogue.
Distribution channels
- PyPI:
pip install meok-mcp-injection-scan-mcp(this package) - Apify Store (Pay-Per-Event): https://apify.com/knowing_yucca/meok-mcp-injection-scan
- GitHub (source): https://github.com/CSOAI-ORG/MEOK-LABS/tree/main/mcps/meok-mcp-injection-scan-mcp
- Sponsor: https://github.com/sponsors/CSOAI-ORG · Pro £79/mo →
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.