product-self-knowledge
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill contains only instructional text and official documentation links. All analyzed content consists of plain-text instructions for routing queries and verifying product details.- [EXTERNAL_DOWNLOADS]: The skill references official documentation sites including docs.claude.com, docs.anthropic.com, support.claude.com, and www.anthropic.com. These are official domains from a well-known service. It also references an official npm package (@anthropic-ai/claude-code) from a trusted organization. No automated download or execution commands are present.- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read and summarize content from the provided documentation URLs. While this represents a data ingestion surface, the targets are restricted to official, trusted product documentation, which minimizes the risk of adversarial content.
Audit Metadata