sparc-refine

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The instruction set is focused on legitimate software engineering tasks, including code review, test coverage improvement, performance profiling, and documentation generation.
  • [COMMAND_EXECUTION]: The skill specifies the use of the Bash tool to execute unit and integration tests, measure code coverage, and profile performance. These operations are restricted to validating the code implementation against provided specifications.
  • [DATA_EXFILTRATION]: Development artifacts and phase states are managed using the mcp__claude-flow__memory toolset. This usage is confined to tracking project progress and maintaining context within the agent's local environment.
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 02:25 AM