tech-talk-outline
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill performs its stated purpose of assisting with conference talk structuring using instructional guidance and established speaking methodologies (e.g., Holman, de Montebello, Heath & Heath).
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-provided content (conference abstracts). While this represents an ingestion point, the skill lacks any dangerous capabilities—such as file system writes outside its context, network operations, or subprocess execution—rendering the risk negligible as it only generates text-based outlines.
- [DATA_EXPOSURE]: The skill references a common project context file (
devrel-context.md) for persistent memory, which is standard practice in its intended environment and does not involve unauthorized access to sensitive system files.
Audit Metadata