documentation-lookup
Pass
Audited by Gen Agent Trust Hub on Apr 11, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill incorporates a security best practice by explicitly instructing the agent to redact secrets, such as API keys and tokens, from queries before they are transmitted to the external Context7 documentation service.
- [PROMPT_INJECTION]: The skill's architecture is susceptible to indirect prompt injection as it ingests untrusted data from an external documentation source (Context7) and processes it within the agent's context.
- Ingestion points: Documentation text and code snippets retrieved via the
query-docsMCP tool inSKILL.md. - Boundary markers: Absent. The skill does not provide instructions to wrap the external content in delimiters or warn the agent to ignore instructions embedded within the fetched documentation.
- Capability inventory: While the skill itself does not define high-risk capabilities like arbitrary command execution or file system writes, it is designed for use in environments like Claude Code or Cursor where such tools may be available to the agent.
- Sanitization: Absent. There is no requirement for the agent to sanitize or validate the content returned from the external documentation service before use.
Audit Metadata