stack-audit
Fail
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: HIGHCOMMAND_EXECUTIONPROMPT_INJECTIONREMOTE_CODE_EXECUTION
Full Analysis
- [COMMAND_EXECUTION] (HIGH): The skill requests broad Bash(bun *) and Bash(rg *) permissions, which allows the agent to execute arbitrary shell commands via these tools.
- [PROMPT_INJECTION] (HIGH): Indirect Prompt Injection risk from untrusted code. 1. Ingestion points: project-root files read via ripgrep and init-audit.ts. 2. Boundary markers: None identified. 3. Capability inventory: Bash shell access, Read file operations, and Glob filesystem navigation. 4. Sanitization: No sanitization of external content before processing.
- [REMOTE_CODE_EXECUTION] (MEDIUM): Executes a local TypeScript file init-audit.ts via the Bun runtime. While ostensibly for auditing, it represents unverified dynamic code execution that performs automated filesystem writes.
Recommendations
- AI detected serious security threats
Audit Metadata