fable-ruki-agenty

Pass

Audited by Gen Agent Trust Hub on Jul 2, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill utilizes GitHub issue bodies as the primary source of instructions and task specifications for sub-agents. This design uses the GitHub CLI to retrieve specifications from a well-known service.
  • Ingestion points: Task specifications are retrieved from GitHub issues via the gh issue view command.
  • Boundary markers: The instructions direct sub-agents to follow specific sections within the issue body such as Goal, Context, Steps, and Definition of Done (DoD).
  • Capability inventory: The workflow involves file system modifications and execution of verification commands by sub-agents.
  • Sanitization: Task content is intended to be generated or updated by the primary orchestration model using gh issue edit.
  • [COMMAND_EXECUTION]: The skill defines a process for sub-agents to run automated verification commands listed in the 'DoD' section of GitHub issues. The workflow incorporates security best practices by requiring a fresh-context verifier agent that is separate from the agent that implemented the task.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 2, 2026, 05:34 AM
Security Audit — agent-trust-hub — fable-ruki-agenty