experiment-design
Pass
Audited by Gen Agent Trust Hub on Apr 20, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests existing project configuration files (YAML, JSON, TOML) to generate matching configuration stubs and execution scripts. This represents a surface where maliciously crafted local files could influence the agent's output. * Ingestion points: Reads existing configuration files in Step 5. * Boundary markers: None present; the instructions do not specify delimiters for external config content. * Capability inventory: Generation of shell scripts and batch runners in Step 6. * Sanitization: None present; no validation or escaping of ingested configuration data is specified.
- [DYNAMIC_EXECUTION]: The skill generates shell scripts and batch runners intended for execution by the user, assembled from templates and project-specific patterns.
Audit Metadata