
Detects hidden AI prompt injection, AI recommendation poisoning, sneaky redirects and hidden spam.
Public tool metadata for what this MCP can expose to an agent.
geo_scan_urlFetches one URL (SSRF-protected, no JS execution), reconstructs redirects, runs all rules.Fetches one URL (SSRF-protected, no JS execution), reconstructs redirects, runs all rules.
No parameter schema in public metadata yet.
geo_trace_redirectsHop-by-hop redirect chain (3xx + Refresh), each hop re-validated; reports declared meta/JS redirects on the final page.Hop-by-hop redirect chain (3xx + Refresh), each hop re-validated; reports declared meta/JS redirects on the final page.
No parameter schema in public metadata yet.
geo_get_reportFull stored report as markdown or json.Full stored report as markdown or json.
No parameter schema in public metadata yet.
Sponsors
GEO_SHIELD_DB_PATHSQLite file for stored scans (default ~/.geo-shield/geo-shield.db).
GEO_SHIELD_RETENTION_DAYSDays to keep stored scans (default 30, 0 = forever).
GEO_SHIELD_ALLOWED_PORTSExtra destination ports to allow besides 80 and 443, comma separated.
A defensive, read-only MCP server that lets AI agents (Claude, Gemini CLI, Cursor, VS Code Copilot, Codex and others) inspect websites for manipulation aimed at search engines and AI assistants: sneaky and open redirects, hidden text and links, injected spam, hijacked metadata, hidden prompt-injection text, and "Summarise with AI" links whose prefilled prompts try to plant instructions in a visitor's assistant ("AI recommendation poisoning").
Every finding is produced by a deterministic rule and backed by evidence: location, visibility state, an excerpt, severity and confidence (kept separate), remediation, and false-positive notes. Scanned content is always treated as untrusted data.
Status: Milestone 1.
geo_scan_url,geo_trace_redirectsandgeo_get_reportare implemented and tested. See Pending for later phases.
geo-shield-mcp