live-cursors
Pass
Audited by Gen Agent Trust Hub on Aug 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists of markdown documentation and architectural guidance. No malicious commands or instructions were found.
- [EXTERNAL_DOWNLOADS]: The skill references downloads from official vendor sources, such as the Rivet container runner binary from
releases.rivet.dev. These are documented as part of the framework's installation process. - [DATA_EXFILTRATION]: No evidence of sensitive data harvesting or unauthorized network transmission. Documentation properly distinguishes between public (
pk_) and secret (sk_) tokens and provides guidance on safe credential management. - [PROMPT_INJECTION]: No attempts to override agent behavior or bypass safety filters were detected in the instructions or metadata.
- [COMMAND_EXECUTION]: While the skill documents the
!commandsyntax supported by some platforms for dynamic context injection, it does so for educational purposes (e.g., teaching developers how the runtime works) and does not use it to execute unauthorized commands.
Audit Metadata