skill-as-a-service
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- Prompt Injection (LOW): Indirect prompt injection surface identified. (1) Ingestion points: The prompt and files parameters in scripts/rebyte_client.py allow untrusted data to enter the execution context. (2) Boundary markers: Absent; the client does not use delimiters or instructions to ignore embedded commands. (3) Capability inventory: The API spawns cloud tasks with access to various skills and file operations. (4) Sanitization: Absent; no input validation is performed before transmission.
- Command Execution (SAFE): Documented shell commands for API interactions and script execution are safe, transparent, and necessary for the skill's operation.
- External Downloads (SAFE): Network requests to api.rebyte.ai and storage.googleapis.com are required for the skill's core functionality and are handled securely via HTTPS.
Audit Metadata