moai-workflow-design-context

Pass

Audited by Gen Agent Trust Hub on Apr 30, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill implements a data ingestion workflow that reads local files and interpolates their content into the agent's context, creating a potential surface for indirect prompt injection.\n
  • Ingestion points: Reads design-related markdown files from .moai/design/ and configuration from .moai/config/sections/design.yaml using the Read tool.\n
  • Boundary markers: Utilizes a specific header ## Design Context (from .moai/design/) and source citations > source: .moai/design/<filename> to delimit injected content, helping the agent distinguish between instructions and data.\n
  • Capability inventory: The skill is restricted to Read, Grep, and Glob tools; it possesses no network access or file-writing capabilities, which significantly limits the potential for exploitation.\n
  • Sanitization: The skill does not perform content sanitization or instruction-filtering on the loaded file content beyond identifying scaffolded _TBD_ markers.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 30, 2026, 12:18 AM