video-color-grading
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides instructions for professional video color grading via the EachLabs API, using standard communication patterns and vendor-controlled infrastructure.
- [CREDENTIALS_UNSAFE]: The documentation follows security best practices by instructing users to provide authentication via an environment variable (
$EACHLABS_API_KEY) rather than hardcoding credentials in scripts or examples. - [DATA_EXFILTRATION]: Network operations are directed towards the vendor's official API and storage domains (
eachsense-agent.core.eachlabs.runandstorage.eachlabs.ai). These operations are consistent with the skill's stated purpose of processing and storing video content. - [INDIRECT_PROMPT_INJECTION]: While the skill defines an interface that processes external video URLs, this is a core requirement for its functional purpose. The documentation focuses on technical parameters for color grading and does not exhibit patterns associated with prompt injection or unsafe data handling.
Audit Metadata