docs-applies-to-tagging
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill fetches documentation guidelines and reference materials from official Elastic domains (elastic.co, elastic.dev). These are trusted sources according to the vendor context and instructions.
- [SAFE]: The skill includes explicit instructions for the agent to only modify files when specifically requested by the user, providing a safety guardrail against unauthorized edits.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied file content and text inputs which could contain malicious instructions. The following analysis of the attack surface confirms the risk is minimal: 1) Ingestion points: Documentation files read via Read, Grep, and Glob tools, as well as text provided in the user prompt; 2) Boundary markers: The skill enforces strict syntax rules for applies_to tags within YAML frontmatter and Markdown blocks; 3) Capability inventory: The skill utilizes Edit for file modifications, WebFetch for external reference checks, and CallMcpTool; 4) Sanitization: The skill performs structural and semantic validation of documentation tags against a predefined schema.
Audit Metadata