python-scratchpad
Pass
Audited by Gen Agent Trust Hub on Mar 21, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
execute_codetool to run Python scripts for arithmetic, unit conversions, and regex validation. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it is designed to ingest and process potentially untrusted external data such as JSON, CSV, or text payloads.
- Ingestion points: Data payloads described in the
SKILL.mdworkflow for transformation tasks. - Boundary markers: No explicit markers or delimiters are defined to separate untrusted data from the agent's instructions.
- Capability inventory: Access to the
execute_codetool is required and explicitly configured in the skill metadata. - Sanitization: The instructions do not specify any validation or sanitization steps for the data being processed by the scripts.
- [DYNAMIC_EXECUTION]: The skill generates and executes Python code at runtime to fulfill task-specific logic, which is the primary intended function of a scratchpad skill.
Audit Metadata