spec
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow that ingests natural language requirements from users to generate EPIC, Story, and Subtask documentation, creating an attack surface for indirect prompt injection.
- Ingestion points: The Requirements Discovery phase in
SKILL.md(Phase 1) is designed to process user-provided natural language inputs to define personas, goals, and values. - Boundary markers: The workflow utilizes a phased approach with mandatory user confirmation commands (e.g., "次に進んで") that provide a human-in-the-loop review step before transitioning to new phases or generating output.
- Capability inventory: The skill is capable of writing multiple markdown files to the local
specs/directory and can invoke the/branchand/prskills to automate Git-related actions based on the generated content as documented inSKILL.md. - Sanitization: No explicit sanitization or filtering mechanisms are defined in the instructions to handle potentially malicious instructions embedded in user-provided requirements before they are interpolated into the final specification documents.
Audit Metadata