causal-loop-diagram-generator

Pass

Audited by Gen Agent Trust Hub on Jul 27, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill's rendering components (cld-inline.js and main.html) fetch the vis-network library from a well-known service (unpkg.com) at a pinned version (10.0.1). This is a common and appropriate method for including visualization dependencies in web-based documentation.
  • [DATA_EXFILTRATION]: The skill includes a 'Save Positions' feature in cld-viewer.js that allow users to download updated configuration or copy it to the clipboard. These actions are performed locally within the browser context to facilitate user-driven layout adjustments; no unauthorized external transmissions were identified.
  • [COMMAND_EXECUTION]: The skill instructions focus on generating JSON and Markdown files and updating the mkdocs.yml configuration. These operations are consistent with the skill's primary purpose of automating documentation tasks and do not involve high-risk shell commands or privilege escalation.
  • [PROMPT_INJECTION]: The skill instructions are clearly defined and focused on diagram generation. There are no attempts to override agent behavior, bypass safety filters, or extract system prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 27, 2026, 01:28 PM
Security Audit — agent-trust-hub — causal-loop-diagram-generator