develop-b2b-golf-feature
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [SAFE]: The skill instructions and metadata do not contain any malicious patterns, obfuscation, or unauthorized access attempts. The inclusion of ASI (AI Security Infrastructure) guardrails demonstrates a strong security posture.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content such as MDX articles and page metadata, creating an attack surface for indirect prompt injection. 1. Ingestion points: src/pages/, src/components/, and MDX article content. 2. Boundary markers: None specified in the instructions. 3. Capability inventory: read_file, write_file, edit_file, create_file, search_code, astro_build, astro_check, astro_dev, and run_tests. 4. Sanitization: The skill mandates strict schema validation for external payloads and form data as per the ASI05 RCE Guard rule.\n- [COMMAND_EXECUTION]: The skill uses a local script (node generate_galleries.mjs) for asset management. This is a project-internal script and its execution is restricted to the local environment and part of the intended developer workflow.
Audit Metadata