agile-product-owner
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes a local Python script (
scripts/user_story_generator.py) to automate the generation of user stories and sprint planning tasks. The script uses only standard library modules and performs safe string formatting and logic. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-provided epic descriptions and requirements to generate structured output. While this creates a surface for indirect prompt injection, the risk is negligible as the skill lacks network access and its output is intended for human review or documentation templates.
- Ingestion points: User-provided requirements and epic details processed via the
user_story_generator.pyscript and markdown templates inSKILL.md. - Boundary markers: The skill uses structured Markdown templates (e.g., Given-When-Then, INVEST checklist) to organize and delimit content.
- Capability inventory: Execution of a local Python script for text generation. No network communication or sensitive file system write capabilities are present.
- Sanitization: The Python script performs standard string cleaning (splitting and case conversion) but does not implement specific LLM prompt sanitization logic.
Audit Metadata