blueprint
Pass
Audited by Gen Agent Trust Hub on Apr 7, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The skill processes untrusted data from brief.md and the codebase to generate implementation plans. This creates a risk of indirect prompt injection where malicious instructions in the brief could influence agent behavior during planning or implementation. * Ingestion points: Content is read from brief.md and the target codebase to fill the plan template. * Boundary markers: The instructions lack delimiters or warnings to ignore instructions embedded within the processed brief. * Capability inventory: The skill performs file writes to create blueprint.md and uses guide-execution to manage slice status and transition into repository implementation work. * Sanitization: There is no evidence of sanitization or validation of the input data before it is used to generate the blueprint.
- [NO_CODE]: The skill does not include any executable scripts, code, or binaries; it consists of a workflow definition and a Markdown template.
Audit Metadata