plotloom-video-adapter

Pass

Audited by Gen Agent Trust Hub on Apr 30, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill is designed to interact with the host environment by executing local CLI tools such as dreamina and plotloom. These tools are used for video submission, status polling, and asset management.
  • [DATA_INGESTION]: The skill reads from local Markdown prompt files (video-prompts.md) to extract information used in CLI commands. This represents a typical data ingestion surface for this type of utility, with the skill providing specific rules for how to translate this content into command flags.
  • [SAFE]: The instructions contain clear security constraints, specifically directing the agent not to automate OAuth flows or store sensitive tokens, which mitigates credential theft risks.
  • [SAFE]: The use of a 'fake' adapter for testing ensures that file path contracts and basic workflow logic can be verified without requiring external network access or consuming service quotas.
  • [SAFE]: Operational transparency is maintained by requiring that task identifiers and status information be recorded in human-readable Markdown notes within the project structure rather than in hidden files.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 30, 2026, 06:25 PM
Security Audit — agent-trust-hub — plotloom-video-adapter