content-strategy
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes a local Python script
scripts/topic_cluster_mapper.pyto perform keyword clustering. This is a standard utility function with no dangerous side effects. - [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided keyword lists (
keywords.txt). While this is an external data source, the processing is limited to lexical clustering via a static script, which presents a negligible attack surface. - [DATA_EXPOSURE]: The skill reads from
.claude/product-marketing-context.mdto gather business context. This is a standard practice for maintaining relevant agent behavior and does not involve sensitive system files or credentials.
Audit Metadata