implement-spec
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from specifications and issue trackers to coordinate implementation tasks.
- Ingestion points: Reads specifications, tickets, and issue tracker content (
SKILL.md). - Boundary markers: None defined; the instructions do not specify delimiters or warnings to ignore instructions embedded within the spec/ticket data.
- Capability inventory: The skill coordinates subagents for exploration and implementation, invokes tools for Test-Driven Development (
tdd) and code review (code-review), and performs git operations like branch creation and Pull Request management (SKILL.md). - Sanitization: No explicit sanitization or validation of the input ticket content is described before it is used to drive agent actions.
Audit Metadata