paw-ps-service-executor
Pass
Audited by Gen Agent Trust Hub on Apr 3, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: Indirect Prompt Injection Surface. The skill ingests data from local project files (e.g., product-context.md, audience-intelligence.md) to inform its service design logic. This ingestion point allows external content to influence agent behavior if the files contain malicious instructions.
- Ingestion points: Reads from .pawbytes/prodig-suites/memory/paw-ps-sidecar/curated/product-context.md, audience-intelligence.md, and .pawbytes/config/config.yaml.
- Boundary markers: Absent. The skill does not use delimiters or specific instructions to ignore embedded commands within the processed files.
- Capability inventory: File system write access to the .pawbytes/prodig-suites/artifacts/ directory and daily logs.
- Sanitization: Absent. Content from the ingested files is interpolated directly into response templates without validation.
Audit Metadata