chinese-copyright-application

Pass

Audited by Gen Agent Trust Hub on Jul 29, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill performs legitimate file analysis on user-provided project directories to extract metadata from configuration files (e.g., package.json, app.json) and README documents.
  • [SAFE]: The provided Python script scripts/generate_copyright_docs.py uses standard system libraries to process files and does not contain any network communication, remote downloads, or obfuscated code.
  • [SAFE]: File system access is appropriately scoped to the user-specified project path and a dedicated output folder. The script explicitly excludes sensitive directories such as .git and node_modules during its scan.
  • [SAFE]: The skill's workflow includes human-in-the-loop checkpoints, such as asking the user for copyright owner details and requiring manual input for sensitive registration information.
  • [SAFE]: While the skill ingests untrusted data from project files (a potential indirect prompt injection surface), the primary extraction logic is handled by a static Python script that formats text rather than executing it. No malicious instructions were found in the templates or processed data examples.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 29, 2026, 07:15 AM
Security Audit — agent-trust-hub — chinese-copyright-application