product0-framing-intent
Pass
Audited by Gen Agent Trust Hub on Jul 13, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill contains no executable scripts, binaries, or remote dependencies. It is composed entirely of natural language instructions and a JSON evaluation file.
- [SAFE]: No patterns of prompt injection, obfuscation, or persistence mechanisms were detected in the instructions or metadata.
- [SAFE]: The skill requests file read/write access to maintain a product brief. No unauthorized access to sensitive credentials, system configurations, or external network domains was identified.
- [NO_CODE]: As the skill does not include any source code, it carries no risk of malicious script execution or unsafe library usage.
- [PROMPT_INJECTION]: A surface for indirect prompt injection exists because the skill ingests raw user input (ideas, complaints) and instructs the agent to store it in a project brief file.
- Ingestion points: Raw requests from users as defined in
SKILL.md. - Boundary markers: None specified for the user content written to the brief.
- Capability inventory: Project file read/write access.
- Sanitization: No escaping or validation is specified for the captured content.
Audit Metadata