product-manager
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes a local shell script,
scripts/init-product.sh, to create a structured directory hierarchy and template files for the product workspace. The script is static, idempotent, and performs routine operations likemkdirandcat. - [INDIRECT_PROMPT_INJECTION]: The skill acts as a data ingestion surface by triaging customer feedback and research insights from external sources.
- Ingestion points: Reads from
discovery/feedback-log.md,discovery/sources.md, anddiscovery/insights.md. - Boundary markers: The instructions mandate strict citation using ID tags (e.g.,
[S#],FB#) and require verbatim copying of external text to prevent interpretive drift or accidental command execution. - Capability inventory: Performs filesystem writes to workspace files and executes the initialization script via
bash. - Sanitization: Includes an adversarial "Judge audit" process to verify the provenance and integrity of data before artifacts are finalized.
Audit Metadata