marketing-ideas
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions direct the agent to read context from a local file,
.claude/product-marketing-context.md, before interacting with the user. This creates an ingestion point for potentially untrusted data that could influence agent behavior if the file contains embedded instructions. - Ingestion points: The file
.claude/product-marketing-context.mdis specified as a context source inSKILL.md. - Boundary markers: The skill lacks explicit instructions to delimit the input or to ignore instructions embedded within the context file.
- Capability inventory: The skill does not possess high-risk capabilities such as network access, file-system modification, or arbitrary command execution.
- Sanitization: There is no evidence of content validation or escaping for the data ingested from the context file.
Audit Metadata