market-patent-diligence
Pass
Audited by Gen Agent Trust Hub on Aug 28, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill uses Python scripts (
scripts/init_diligence_pack.pyandscripts/validate_diligence_pack.py) to automate the creation and validation of markdown-based diligence reports. These scripts use standard library functions likepathlibfor file operations andargparsefor user input, restricted to local directory management. - [NO_CODE]: The core functionality of the skill is delivered through instructional markdown and reference guides, with scripts serving only as structural helpers.
- [REMOTE_CODE_EXECUTION]: The skill explicitly instructs the agent to search official and public sources (e.g., USPTO, WIPO, SEC filings) and manually log findings. It does not perform automated scraping or execute code fetched from the internet.
- [PRIVILEGE_ESCALATION]: No privilege escalation patterns detected. Scripts run with standard user permissions to write local files.
- [DATA_EXFILTRATION]: The skill focuses on aggregating public information into a local directory ('diligence pack'). It does not contain network exfiltration patterns or sensitive data access.
Audit Metadata