
firmis-scanner
by Firmislabs·★ 0·Score 36
Firmis is a security scanner that detects malicious behavior in Claude Skills, MCP Servers, and other AI agent platforms.
Overview
Firmis provides runtime security scanning for AI agents, protecting against credential harvesting, prompt injection, tool poisoning, and 18 other threat categories. It scans both code surfaces (what the agent's code actually does) and instruction surfaces (what SKILL.md and tool descriptions tell the agent to do). The tool can be used as a standalone scanner, integrated with Claude Desktop via MCP, or embedded in CI/CD pipelines with SARIF output.
Try asking AI
After installing, here are 5 things you can ask your AI assistant:
When to choose this
Choose Firmis when you need security scanning for multiple AI agent platforms (MCP, Claude Skills, Codex plugins, etc.) and want deterministic rule-based analysis without requiring AI model inference.
When NOT to choose this
Don't choose Firmis if you need runtime monitoring of already-deployed agents (it focuses on pre-deployment scanning) or if you require AI-powered exploit analysis (which is a paid feature).
Tools this server exposes
3 tools extracted from the READMEfirmis_scanScan AI agent projects for security threats
firmis_discoverDiscover supported platforms and configurations
firmis_reportGenerate security reports
Comparable tools
Installation
Installation
- **Zero-install scan**:
``bash npx firmis-cli scan ``
- **Global installation**:
``bash npm install -g firmis-cli firmis scan ``
Claude Desktop Integration
Add to your MCP configuration:
{
"mcpServers": {
"firmis": {
"command": "npx",
"args": ["-y", "firmis-cli", "--mcp"]
}
}
}This provides firmis_scan, firmis_discover, and firmis_report tools.
FAQ
- Is it free?
- Yes. The scanner is free, open-source (Apache-2.0), and requires no account. Run `npx firmis-cli scan` — unlimited scans, all rules, HTML + JSON + SARIF reports.
- What is tool poisoning?
- Tool poisoning is when an MCP server embeds hidden instructions in tool descriptions to hijack your AI agent. Research shows a 72.8% attack success rate. Firmis detects known poisoning patterns, hidden Unicode, description/behavior mismatches, and prompt override attempts.
Compare firmis-scanner with
Last updated · Auto-generated from public README + GitHub signals.