skill-with-scripts
Fail
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: HIGHCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION] (HIGH): The skill instructions (SKILL.md) direct the agent to execute shell commands like
python scripts/process_data.py --input <file>andpython scripts/validate.py --data <input>. Direct interpolation of user-controlled variables into these command strings poses a severe risk of command injection if the underlying agent implementation does not perform rigorous shell escaping. - [PROMPT_INJECTION] (HIGH): (Category 8: Indirect Prompt Injection) The skill is designed to ingest and process untrusted external data with high-privilege execution capabilities.
- Ingestion points: Untrusted data is ingested via the
--input(line 35) and--data(line 41) parameters. - Boundary markers: Absent. The instructions do not define delimiters or provide 'ignore instructions' warnings for the content being processed.
- Capability inventory: The skill possesses the capability to execute local Python scripts, which can perform arbitrary file system or system operations depending on the script logic.
- Sanitization: While a
validate.pyscript is mentioned, there is no evidence of strict schema validation or sanitization of input before it is used to construct the execution command.
Recommendations
- AI detected serious security threats
Audit Metadata