plan-implementation
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from issue titles and descriptions to generate implementation plans.\n
- Ingestion points: External issue content is read from the issue tracker or provided via user context (SKILL.md).\n
- Boundary markers: The instructions lack specific delimiters or 'ignore' directives to separate untrusted issue data from the agent's instructions.\n
- Capability inventory: The skill has capabilities to write markdown files (PLAN.md, CONTEXT.md) and update local issue state files (SKILL.md).\n
- Sanitization: There is no explicit requirement to sanitize or validate the issue content before it is incorporated into the generated plan.
Audit Metadata