clay-toolkit
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill includes a Python script
scripts/check-output.pywhich is used for validating that generated deliverables contain required Markdown sections. The script uses standard libraries (pathlib,sys) and performs simple string matching on file contents. It does not execute arbitrary shell commands or access network resources. - [DATA_EXPOSURE]: The skill contains JavaScript-like formulas intended for the Clay platform (e.g., domain extraction, ICP scoring). These are standard data transformation patterns and do not involve sensitive credential handling or exfiltration techniques.
- [METADATA_POISONING]: The metadata fields in
SKILL.md(author, version, tool_path) are consistent with the skill's stated purpose of providing GTM (Go-To-Market) tooling. The vendor 'LeadMagic' is identified as the author, and all references to LeadMagic tools and patterns are consistent with this identity. - [INDIRECT_PROMPT_INJECTION]: While the skill defines blueprints for AI-driven enrichment (e.g., Blueprint 1 in
gtm-table-blueprints.md), it functions as a template provider rather than an active processor of untrusted external data. The capability surface is limited to structural design and formula generation.
Audit Metadata