{
 "public_id": "AIR-2026-0242",
 "incident_id": "2026-09-praisonai-framework-auth-sandbox-cluster",
 "date_disclosed": "2026-09-14",
 "date_occurred": null,
 "title": "PraisonAI multi-agent framework: coordinated September 2026 disclosure of unauthenticated tool/agent execution and sandbox escapes",
 "class": "responsible_disclosure",
 "attack_surface": "agent_framework",
 "initial_vector": "unauthenticated_network_request",
 "agent_capability_abused": [
  "tool_invocation",
  "code_execution",
  "agent_job_submission",
  "mcp_stdio_spawn",
  "approval_gate"
 ],
 "target_product": "PraisonAI (npm praisonai / praisonai-ts and pip praisonai, praisonaiagents)",
 "target_vendor": "MervinPraison",
 "autonomy_level": "semi_autonomous",
 "impact_realized": false,
 "impact_type": [
  "remote_code_execution",
  "sandbox_escape",
  "authorization_bypass",
  "unauthorized_tool_execution"
 ],
 "blast_radius": "PraisonAI < 4.6.59 / praisonaiagents < 1.6.59 and praisonai-ts codeMode/MCPServer 1.4.0-1.7.1; HTTP/SSE/Jobs surfaces bind 0.0.0.0 by default",
 "cve": [
  "CVE-2026-57138",
  "CVE-2026-57139",
  "CVE-2026-57123",
  "CVE-2026-57124",
  "CVE-2026-57125",
  "CVE-2026-57127",
  "CVE-2026-57131"
 ],
 "taxonomy": {
  "owasp_asi": [
   "ASI05",
   "ASI03"
  ],
  "mitre_atlas": [],
  "labels": [
   "missing_authentication",
   "sandbox_escape",
   "approval_bypass",
   "dns_rebinding",
   "agent_framework",
   "umbrella"
  ]
 },
 "defense_present": "bundled MCPSecurity / SecurityConfig / @require_approval and codeMode source blocklist (none wired into the vulnerable paths)",
 "defense_bypassed": true,
 "fix_shipped": "praisonai 4.6.58/4.6.59, praisonaiagents 1.6.59, praisonai-ts 1.7.2",
 "primary_sources": [
  "https://cveawg.mitre.org/api/cve/CVE-2026-57139",
  "https://github.com/MervinPraison/PraisonAI/security/advisories/GHSA-j4f3-55x4-r6q2",
  "https://github.com/MervinPraison/PraisonAI/security/advisories/GHSA-vmmj-pfw7-fjwp",
  "https://github.com/MervinPraison/PraisonAI/security/advisories/GHSA-4869-x4pr-q22x",
  "https://cvebrief.com/archive/2026/09/16/"
 ],
 "verified_quote": "MCPServer.startHttp() in src/praisonai-ts/src/mcp/server.ts binds without a host restriction and forwards every HTTP POST request to handleRequest() without authentication or authorization.",
 "confidence": "high",
 "notes": "UMBRELLA record for the coordinated Sept 2026 PraisonAI disclosure batch (all CVSS 9.8-9.9, published 2026-09-14/15). Fused per the corpus umbrella precedent (cf. AIR-2026-0089 'MCP by Design' bundling a dozen CVEs) because they share one framework and root theme: security controls exist in the codebase but are never wired into the HTTP/SSE/MCP/Jobs execution paths, plus a codeMode 'sandbox' that is only a source blocklist. Mechanisms: 57138 codeMode Function-constructor escape; 57139 MCPServer unauth tools/call; 57123 SSE bind 0.0.0.0 no auth/origin; 57124 /api/mcp/connect unauth local command; 57125 Jobs API unauth RCE + approve-field approval bypass; 57127 auth middleware silently disabled when no secret set; 57131 Jobs API no auth. Reporters include rexpository, vndasunkid and others. Related same-batch CVE-2026-57122 (webhook HMAC bypass, 8.6) and CVE-2026-57124's twin 57123 also disclosed. All responsible disclosure, PoC-level, no real victim.",
 "risk_category": "security_fraud",
 "risk_secondary": [],
 "risk_subcategory": "Server-Side Request Forgery via LLM",
 "risk_sub_secondary": [
  "Remote Code Execution & Sandbox Escape",
  "Unauthorized Access & Account Takeover",
  "Multi-agent contagion"
 ],
 "ai_entity": "generative",
 "agency": "ai_as_target",
 "agency_debatable": false,
 "agency_reason": "framework authentication and sandbox flaws; no model decision",
 "vector_family": "conventional_vulnerability",
 "vector_detail": "unauthenticated_network_request",
 "guardrail_kind": [
  "human_approval",
  "permission_scope"
 ],
 "guardrail_outcome": "silent",
 "xrefs": {
  "aiid": [],
  "atlas": [],
  "cve": [
   "CVE-2026-57138",
   "CVE-2026-57139",
   "CVE-2026-57123",
   "CVE-2026-57124",
   "CVE-2026-57125",
   "CVE-2026-57127",
   "CVE-2026-57131"
  ],
  "ghsa": []
 },
 "primary_asi": "ASI05",
 "agent_framework": "unknown",
 "model_provider": "unknown",
 "model_name": "",
 "detection_mechanism": "unknown",
 "tool_access": "unknown",
 "financial_loss_band": "unknown",
 "eai_category": "confused_deputy",
 "eai_domain": "access_control_and_permissions",
 "eai_secondary": [
  "privilege_escalation"
 ],
 "eai_reason": "subcategory: Server-Side Request Forgery via LLM",
 "asi_curated": [],
 "asi_scope": "in_scope",
 "asi_provenance": "rule",
 "asi_reason": "ASI05 (impact=code_exec); ASI03 (eai_category=confused_deputy)",
 "submitted_by": "EAI Internal"
}