code-explain
Fail
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
- Remote Code Execution (HIGH): The skill relies on 'npx ai-explain', which fetches and runs a package from the public npm registry. Since the author and package are not from a trusted source, this constitutes a high-risk RCE vector.
- Indirect Prompt Injection (HIGH): 1. Ingestion points: The skill reads arbitrary code files (e.g., ./src/utils/crypto.ts) provided as input. 2. Boundary markers: Absent. There are no instructions to treat content as untrusted data. 3. Capability inventory: The skill executes shell commands via npx. 4. Sanitization: None detected. External content is passed directly to the execution tool.
- Credentials Unsafe (MEDIUM): The skill explicitly requires an OPENAI_API_KEY. Malicious code executed via the npx package could access and exfiltrate this key from the environment.
- External Downloads (MEDIUM): The skill fetches the 'ai-explain' package from an unverified source (LXGIC Studios) rather than a trusted organization.
Recommendations
- AI detected serious security threats
Audit Metadata