knowledge-base-for-startups
Pass
Audited by Gen Agent Trust Hub on Nov 17, 2026
Risk Level: SAFENO_CODE
Full Analysis
- Factual Data Handling: The skill is configured to provide factual answers based on bundled reference files. By design, the instructions require the agent to read files before answering and quote them verbatim, which minimizes the risk of the model hallucinating incorrect information.
- Instructional Integrity Guards: The skill instructions include an explicit safety boundary directing the AI agent to ignore any commands or behavioral overrides that might be encountered in the reference files. This is a proactive measure against potential indirect prompt injection considerations.
- Verified External Resources: All external links in the skill point to official vendor domains or verified partner offers. These are documented for the intended purpose of facilitating startup growth and do not involve unauthorized data exfiltration or suspicious redirections.
- Skill Maintenance Procedures: The skill includes a conditional logic check to suggest that the user update their local version if the content is outdated. This ensures the user continues to receive accurate information through established skill management tools.
Audit Metadata