gsp-rolling-supervisor

Pass

Audited by Gen Agent Trust Hub on May 3, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill defines a set of architectural rules and a worker graph for an autonomous build-verify-repair loop, which is a standard orchestration pattern.
  • [PROMPT_INJECTION]: The skill implements a rolling state mechanism where the agent reads from and writes to state files (docs/game-studio/rolling/state.json). This creates a surface for indirect prompt injection as data from previous sessions or the project repository is re-ingested into the agent context.
  • Ingestion points: docs/game-studio/rolling/state.json and project files during the build/verify/repair cycles.
  • Boundary markers: No explicit delimiters or instructions to ignore instructions embedded in the project files are provided.
  • Capability inventory: The skill possesses file-writing capabilities and instructions to execute repository-level build and repair commands.
  • Sanitization: There is no evidence of sanitization or validation of the ingested state or project data before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
May 3, 2026, 01:47 PM
Security Audit — agent-trust-hub — gsp-rolling-supervisor