the-kpi-blueprint
Pass
Audited by Gen Agent Trust Hub on Aug 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill incorporates an exhaustive security reference in
references/agent-security.mdthat teaches the agent to distinguish between data and instructions, effectively neutralizing indirect prompt injection attempts from fetched or pasted content. - [SAFE]: Advanced safety measures against Cross-Site Scripting (XSS) are implemented through mandatory HTML escaping, attribute quoting, and URL validation for all generated code and markup, ensuring the output is safe for deployment.
- [SAFE]: The 'Missing Input Protocol' in
references/missing-input-protocol.mdestablishes a rigorous framework for data transparency, preventing the agent from silently filling information gaps or fabricating metrics when inputs are absent. - [SAFE]: Explicit privacy safeguards are included to prevent the processing of special-category (protected) data and to ensure the redaction of any credentials found in source files or exports.
- [SAFE]: The skill follows the principle of least privilege by using local context files (e.g.,
.agents/product-context.md) and avoiding any external network operations or unverified dependencies.
Audit Metadata