refactor-backend-09-code-optimize

Pass

Audited by Gen Agent Trust Hub on Jul 6, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a local script scripts/checklist.mjs using the Node.js runtime to output a machine-readable JSON checklist of the skill's operational steps for verification purposes.
  • [PROMPT_INJECTION]: The skill processes external backend source code as part of its refactoring workflow, which constitutes an indirect prompt injection surface.
  • Ingestion points: The skill reads files annotated with @Service or @Repository within the user's project codebase as identified in SKILL.md and references/method.md.
  • Boundary markers: There are no explicit instructions or delimiters defined to separate the ingested source code from the agent's internal instructions.
  • Capability inventory: The skill has file-writing capabilities used to perform the refactoring operations described in the 'fix' mode.
  • Sanitization: The skill employs deterministic refactoring logic (SQL-1 through OPT-5) and requires a mandatory 'style checkpoint' where the user must approve the first file modification before the agent continues.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 6, 2026, 01:01 AM
Security Audit — agent-trust-hub — refactor-backend-09-code-optimize