implement-spec

Pass

Audited by Gen Agent Trust Hub on Sep 20, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon external specifications and tickets, which are untrusted data sources.
  • Ingestion points: The instructions in SKILL.md require the agent to read and understand external specifications and task tickets to build a task graph.
  • Boundary markers: The instructions lack explicit delimiters or safety markers to differentiate the data in the specifications from the agent's operational logic.
  • Capability inventory: The skill coordinates subagents with capabilities to write files (exploration notes), manage worktrees, create branches, and merge code.
  • Sanitization: There is no requirement or mechanism mentioned for sanitizing or validating the content of the external specifications before they are used to drive subagent actions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 20, 2026, 07:48 AM
Security Audit — agent-trust-hub — implement-spec