reels-script-writer

Pass

Audited by Gen Agent Trust Hub on Aug 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The repository contains shell scripts (tests/validate_repo.sh, tests/validate_negative_cases.sh) and a Python script (tests/score_eval_results.py) designed for repository integrity verification and performance scoring. These scripts facilitate standard automated testing and development workflows.
  • [EXTERNAL_DOWNLOADS]: The documentation specifies an installation process using npx skills that targets the author's public GitHub repository. This aligns with standard procedures for deploying skills in this environment and involves a known vendor resource.
  • [SAFE]: A proactive security scanner is included in tests/validate_repo.sh which uses regex patterns to check for hardcoded secrets, such as GitHub tokens and AWS keys. This feature serves as a guardrail to prevent the accidental distribution of sensitive credentials.
  • [REMOTE_CODE_EXECUTION]: The manual evaluation instructions in README.md reference a script at a hardcoded local path (/Users/zhengyujie/...). Although this path is specific to a developer's local machine and represents an unverifiable external reference, it is presented as an optional testing step rather than a core component of the skill's execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 20, 2026, 05:24 AM
Security Audit — agent-trust-hub — reels-script-writer