{
 "public_id": "AIR-2025-0029",
 "incident_id": "2025-04-trailofbits-mcp-conversation-theft",
 "date_disclosed": "2025-04-23",
 "date_occurred": null,
 "title": "Malicious MCP tool descriptions can exfiltrate a user's entire conversation history on a trigger word",
 "class": "research_demo",
 "attack_surface": "mcp_server",
 "initial_vector": "malicious_tool_description",
 "agent_capability_abused": [
  "tool_description_ingestion",
  "tool_invocation"
 ],
 "target_product": "MCP clients generally",
 "target_vendor": "multiple",
 "autonomy_level": "copilot_with_approval",
 "impact_realized": false,
 "impact_type": [
  "data_exfiltration"
 ],
 "blast_radius": "Any environment where a malicious MCP server is installed",
 "cve": [],
 "taxonomy": {
  "owasp_asi": [
   "ASI01",
   "ASI04",
   "ASI02"
  ],
  "mitre_atlas": [],
  "labels": [
   "tool_poisoning",
   "mcp",
   "trigger_phrase",
   "persistent_exfiltration"
  ]
 },
 "defense_present": null,
 "defense_bypassed": true,
 "fix_shipped": null,
 "primary_sources": [
  "https://blog.trailofbits.com/2025/04/23/how-mcp-servers-can-steal-your-conversation-history/"
 ],
 "verified_quote": "Conversation histories are a gold mine of sensitive information, and this type of attack can be persistent: unlike traditional point-in-time data breaches, a single, sustained connection to a malicious MCP server could continue to compromise and harvest weeks or months of conversations.",
 "confidence": "high",
 "notes": "Author Keith Hoodlet. Second post in the Trail of Bits MCP line-jumping series.",
 "risk_category": "privacy",
 "risk_secondary": [],
 "risk_subcategory": "Data Exfiltration via Context",
 "risk_sub_secondary": [],
 "ai_entity": "generative",
 "agency": "agent_acted",
 "agency_debatable": false,
 "agency_reason": "the agent exfiltrated the conversation history on a trigger word",
 "vector_family": "malicious_component",
 "vector_detail": "malicious_tool_description",
 "guardrail_kind": [
  "unknown"
 ],
 "guardrail_outcome": "silent",
 "xrefs": {
  "aiid": [],
  "atlas": [],
  "cve": [],
  "ghsa": []
 },
 "primary_asi": "ASI01",
 "agent_framework": "mcp_server",
 "model_provider": "unknown",
 "model_name": "",
 "detection_mechanism": "external_researcher",
 "tool_access": "unknown",
 "financial_loss_band": "unknown",
 "eai_category": "data_exfiltration_channels",
 "eai_domain": "privacy",
 "eai_secondary": [
  "supply_chain_vulnerabilities"
 ],
 "eai_reason": "subcategory: Data Exfiltration via Context",
 "asi_curated": [
  "ASI01"
 ],
 "asi_scope": "in_scope",
 "asi_provenance": "curated_plus_rule",
 "asi_reason": "curated: ASI01 | added ASI04 (vector_family=malicious_component), ASI02 (impact=tool_misuse)",
 "submitted_by": "EAI Internal"
}