docs-seeker

Pass

Audited by Gen Agent Trust Hub on Mar 7, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core function of fetching and processing external technical documentation.\n
  • Ingestion points: External documentation data enters the agent context via fetch-docs.js and is further processed by analyze-llms-txt.js.\n
  • Boundary markers: Absent. The skill does not implement XML tags, delimiters, or specific safety instructions to the agent to treat retrieved documentation as untrusted content.\n
  • Capability inventory: The skill executes internal Node.js scripts and includes workflows that suggest cloning repositories and using CLI tools like repomix.\n
  • Sanitization: No sanitization, filtering, or validation is performed on the content retrieved from external documentation URLs.\n- [EXTERNAL_DOWNLOADS]: Fetches technical documentation and configuration metadata from context7.com and GitHub. These are recognized as well-known technology services and are used for their intended purpose; these interactions are documented neutrally.\n- [COMMAND_EXECUTION]: Executes included scripts (detect-topic.js, fetch-docs.js, analyze-llms-txt.js) using Node.js to classify queries and automate documentation retrieval. The env-loader.js utility implements a hierarchical search for environment configuration, which is a common pattern for managing tool settings.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 7, 2026, 07:27 PM