site-architecture
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read context from specific project files (
.agents/product-marketing.md,.claude/product-marketing.md, orproduct-marketing-context.md) to inform its site architecture recommendations. This creates a surface where an attacker who can modify these files could inject instructions to bias the agent's output. - Ingestion points: Instructions in
SKILL.mdto read product marketing context files. - Boundary markers: Absent; the instructions do not provide delimiters or warnings for the agent to ignore instructions found within these files.
- Capability inventory: The skill is limited to generating text, ASCII trees, and Mermaid diagrams. No file writes, network exfiltration, or shell command execution capabilities are defined.
- Sanitization: None; the skill does not validate the content of the ingested context files.
Audit Metadata