{
  "markdown": "# TrustScan — Scan Before You Trust\n\n> Security scanner for MCP servers and AI skills. Detects invisible Unicode\n> prompt-injection, dangerous code patterns (MCP001–006), hardcoded secrets,\n> and typosquat package names. Trust scoring for the agent economy.\n\n[![Live API](https://img.shields.io/badge/API-live-success)](https://trust-scan-production.up.railway.app/health)\n[![Landing](https://img.shields.io/badge/status-page-blue)](https://trust-scan-production.up.railway.app/status)\n[![MCP Registry](https://img.shields.io/badge/MCP-io.github.entradox%2Ftrust--scan-purple)](https://registry.modelcontextprotocol.io)\n\n## Why\n\nAgents install MCP servers and skills from public registries with no review.\nA single invisible Unicode character in a tool description is enough to inject\ninstructions into the agent running it. TrustScan scans before you trust.\n\n**Proven in production:** TrustScan's own verification run found a real\ninvisible-Unicode injection flag in a sibling product's codebase (38 files scanned).\n\n## What it checks\n\n| Check | What it catches |\n|---|---|\n| Invisible Unicode | Prompt-injection payloads hidden in homoglyph/zero-width chars |\n| Dangerous patterns (MCP001–006) | Dynamic imports, subprocess shells, credential exfil shapes |\n| Secrets | Hardcoded API keys and tokens |\n| Typosquats | Package names one edit-distance from popular packages |\n\n## Try it in 30 seconds — no signup\n\n```bash\ncurl -X POST https://trust-scan-production.up.railway.app/v1/scan \\\n  -H \"Content-Type: application/json\" \\\n  -d '{\"path\": \"/path/to/mcp-server-or-skill\"}'\n```\n\n## Install (MCP clients)\n\n```json\n{\n  \"mcpServers\": {\n    \"trust-scan\": {\n      \"url\": \"https://trust-scan-production.up.railway.app/mcp/\"\n    }\n  }\n}\n```\n\nTools: `trust_scan_server` (directory/package), `trust_scan_file` (single file).\n\nAgent-facing API reference: [`llms.txt`](https://trust-scan-production.up.railway.app/llms.txt)\n\n## Architecture\n\n- Scanner (`trust_scan.py`): severity-weighted findings, Levenshtein typosquat\n  detection, invisible-codepoint analysis\n- REST API (`api_server.py`, FastAPI): `/health`, `/v1/scan`, `/stats`,\n  hosted streamable-http MCP at `/mcp/`\n\n## Contact\n\nentradox@icloud.com",
  "bytes": 2199,
  "sha": "740fd393c6db8767c0167535247c84921ecc07ed923f46f389c96759329e2c06",
  "repo_slug": "entradox/trust-scan",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_github_entradox_trust_scan_12b7c26e/readme"
}