copy-editing
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input in the form of marketing copy for editing and review. It also directs the agent to read local context files (such as
.agents/product-marketing.md) to guide the editing process. This creates a minor vulnerability surface where malicious instructions embedded in the user-provided copy could attempt to influence the agent's behavior or extract information from the marketing context files. However, the skill lacks high-risk capabilities such as network communication, file writing, or shell command execution, which mitigates the risk. - Ingestion points: User-provided marketing copy for editing (SKILL.md).
- Boundary markers: Absent. The skill does not explicitly instruct the agent to use delimiters or warnings to ignore instructions within the processed text.
- Capability inventory: Reading local project files (
.agents/product-marketing.md,.claude/product-marketing.md,product-marketing-context.md) for context. - Sanitization: None identified. The skill assumes the input is standard marketing text to be edited.
Audit Metadata