anthropic-validator
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches current standards from official Anthropic documentation URLs (e.g., docs.anthropic.com). This is documented as safe behavior as it targets a trusted organization to ensure validation rules remain current.
- [PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection because it processes untrusted content from user-provided files during the validation process.
- Ingestion points: Asset files (skills, hooks, agents, etc.) provided by the user are read and passed to sub-agents for analysis.
- Boundary markers: The skill does not implement specific delimiters or instructions to ignore embedded commands within the asset content being analyzed.
- Capability inventory: The skill possesses capabilities for reading local files, writing validation reports to the logs/ directory, and spawning sub-agents via the Task tool.
- Sanitization: No sanitization or validation of the input asset content is performed prior to analysis.
Audit Metadata