sk-diverge

Pass

Audited by Gen Agent Trust Hub on Apr 22, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to check the local file system (specifically workspace/sessions/) to identify active sessions and read specific templates. It also involves writing session results to Markdown files and updating session state YAML files. These operations are scoped to a local workspace environment and represent standard application state management.
  • [DATA_EXFILTRATION]: No network tools (like curl or wget) or external API calls are used. Data processed (user skills, passions, and problems) is stored locally within the workspace and is not transmitted externally.
  • [EXTERNAL_DOWNLOADS]: The skill does not reference or attempt to download any external scripts, binaries, or packages.
  • [REMOTE_CODE_EXECUTION]: There is no evidence of dynamic code generation, use of eval/exec functions, or execution of untrusted remote content.
  • [PROMPT_INJECTION]: The instructions are instructional and pedagogical, focusing on brainstorming methodology. There are no attempts to bypass safety filters or override system-level constraints.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 22, 2026, 01:13 PM