business-plan
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEMETADATA_POISONINGNO_CODE
Full Analysis
- [SAFE]: The skill contains no executable code, shell scripts, or commands. It is purely instructional text for guiding the AI's response format.
- [METADATA_POISONING]: The
authorfield in the YAML frontmatter is set to "Claude Code", which is the name of a specific CLI tool by Anthropic. However, the footer link indicates the skill is part of a third-party collection by "inbharatai". This discrepancy in author attribution is misleading but does not pose a functional security risk to the user environment. - [EXTERNAL_DOWNLOADS]: The skill contains a reference to an external GitHub repository (
github.com/inbharatai/claude-skills). This is a static link in the documentation for informational purposes and does not trigger automated downloads, dependency installation, or remote code execution.
Audit Metadata