writing-plans
Pass
Audited by Gen Agent Trust Hub on Jul 28, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it transforms external specifications into actionable implementation plans.
- Ingestion points: User-provided requirements or specifications processed in SKILL.md.
- Boundary markers: Absent; there are no delimiters or instructions to ignore embedded commands within the input specifications.
- Capability inventory: The generated plans result in file modifications and shell command execution via the agent or subagents.
- Sanitization: Absent; the skill does not provide mechanisms to escape or validate input data before it is included in the plan.
- [COMMAND_EXECUTION]: The implementation plan templates include shell commands for test execution (
pytest) and version control (git), which are intended for execution by the agent or referenced sub-skills.
Audit Metadata