{
 "public_id": "AIR-2025-0153",
 "incident_id": "2025-10-claude-code-home-directory-deletion",
 "date_disclosed": "2025-10-21",
 "date_occurred": "2025-10-21",
 "title": "Claude Code executed a recursive delete from root, destroying all user-owned files in a developer's home directory",
 "class": "safety_failure",
 "attack_surface": "coding_agent",
 "initial_vector": "autonomous_agent_action",
 "agent_capability_abused": [
  "shell_access",
  "filesystem_write"
 ],
 "target_product": "Claude Code",
 "target_vendor": "Anthropic",
 "autonomy_level": "semi_autonomous",
 "impact_realized": true,
 "impact_type": [
  "data_destruction"
 ],
 "blast_radius": "one developer's home directory on Ubuntu/WSL2; weeks-to-months of project work, only dotfiles survived",
 "cve": [],
 "taxonomy": {
  "owasp_asi": [
   "ASI10",
   "ASI02"
  ],
  "mitre_atlas": [],
  "labels": [
   "destructive_tool_use",
   "rm_rf",
   "insufficient_command_guardrails"
  ]
 },
 "defense_present": "OS file permissions (only reason system files survived); user interrupt (too late)",
 "defense_bypassed": true,
 "fix_shipped": null,
 "primary_sources": [
  "https://github.com/anthropics/claude-code/issues/10077"
 ],
 "verified_quote": "Claude Code executed a destructive recursive delete command that successfully deleted all user files in my home directory.",
 "confidence": "high",
 "notes": "Reported by GitHub user mikewolak as issue #10077, titled '[BUG] ## CRITICAL: Claude Code executed rm -rf deleting entire home directory'. The command reportedly targeted / and failed only on permission-protected system paths. No attacker; the agent generated the command while working in a firmware project subdirectory.",
 "risk_category": "reliability",
 "risk_secondary": [],
 "risk_subcategory": "Destructive or irreversible action",
 "risk_sub_secondary": [],
 "ai_entity": "generative",
 "agency": "agent_acted",
 "agency_debatable": false,
 "agency_reason": "Claude Code executed rm -rf and destroyed the home directory",
 "vector_family": "none_autonomous",
 "vector_detail": "autonomous_agent_action",
 "guardrail_kind": [
  "permission_scope"
 ],
 "guardrail_outcome": "silent",
 "xrefs": {
  "aiid": [],
  "atlas": [],
  "cve": [],
  "ghsa": []
 },
 "primary_asi": "ASI10",
 "agent_framework": "claude_code",
 "model_provider": "anthropic",
 "model_name": "",
 "detection_mechanism": "unknown",
 "tool_access": "unknown",
 "financial_loss_band": "unknown",
 "eai_category": "unsafe_actuation",
 "eai_domain": "agent_behaviour",
 "eai_secondary": [],
 "eai_reason": "subcategory: Destructive or irreversible action",
 "asi_curated": [],
 "asi_scope": "in_scope",
 "asi_provenance": "rule",
 "asi_reason": "ASI10 (no adversary + unsafe actuation with no attacker-supplied trigger); ASI02 (impact=tool_misuse)",
 "submitted_by": "EAI Internal"
}