In the wild
Realized harm
QUIETVAULT: credential stealer that invokes on-host AI CLI tools to hunt for additional secrets
AIR-2025-0170 AIR record · 2025-11-gtig-quietvault
A real party was affected. This incident caused actual consequences, not a demonstrated capability.
From the supporting source
QUIETVAULT leverages an AI prompt and on-host installed AI CLI tools to search for other potential secrets on the infected system and exfiltrate these files to GitHub as well.cloud.google.com
The record
Disclosed
2025-11-05
Occurred
2025
Vendor
various
Product
Locally installed AI CLI tools (e.g. Claude Code, Gemini CLI)
Risk domain
Access control and permissions › Credential theft
Also touches
Data exfiltration channels
System type
Generative model
Who acted
Agent acted — the stealer invokes the on-host AI CLIs, which do the secret hunting
Attack surface
Coding agent
Autonomy level
Semi-autonomous
Initial vector
Javascript stealer on developer host
Impact type
Credential theft, Data exfiltration
Confidence
High
Submitted by
EAI Internal
Crosswalk
- OWASP Agentic Top 10 (2026)
- ASI03 – Identity and Privilege Abuse
- MITRE ATLAS
- AML.T0012 – Valid Accounts
- NIST AI RMF
- MEASURE 2.7
- EU AI Act
- Article 13
- ISO/IEC
- TR 24028; 42001; 23894
- AIUC-1
- B. Security
- Scope
- tool failure
- Mitigation
- Use secure credential storage; Implement short-lived tokens.
- Taxonomy
- Enkrypt AI agent risk taxonomy · Black-Box Red Teaming of Agentic AI: A Taxonomy-Driven Framework for Automated Risk Discovery
Blast radius
Developers with GitHub and NPM tokens; stolen data pushed to attacker-created public GitHub repositories
Capability abused
Taxonomy & references
Curator notes
Same 'living off the AI tools already installed' pattern as the Nx s1ngularity payload. GTIG marks it as observed in operations, so impact_realized is true, but no victim count is given.
Supporting sources
Cite as
Enkrypt AI. "QUIETVAULT: credential stealer that invokes on-host AI CLI tools to hunt for additional secrets". AIR, AIR-2025-0170. https://agentincident.vercel.app/i/AIR-2025-0170/ (accessed 2026-10-19).