component-docs-batcher
Fail
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: HIGHCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- PROMPT_INJECTION (HIGH): The skill is highly vulnerable to Indirect Prompt Injection (Category 8). 1. Ingestion points: Component source files (.tsx, .jsx) and structural metadata. 2. Boundary markers: Absent; there are no delimiters or instructions to ignore embedded content. 3. Capability inventory: The agent can execute local scripts via
npx ts-nodeand write/modify markdown files. 4. Sanitization: Absent; the skill relies on 'understanding' logic from raw code. Maliciously crafted comments in source code could manipulate the agent's behavior during documentation or task-tracking. - COMMAND_EXECUTION (HIGH): The skill instructs the agent to run local TypeScript scripts (
scan-components.ts,generate-todos.ts,analyze-component.ts) usingnpx ts-node. This allows for arbitrary code execution on the host system if the skill originates from an untrusted source or if its scripts are compromised. - DATA_EXPOSURE (LOW): The skill accesses the local file system and Git history (
git log) to identify component status and metadata. While functional, this behavior exposes internal project structure and history to the agent context.
Recommendations
- AI detected serious security threats
Audit Metadata