subagent-driven-development

Pass

Audited by Gen Agent Trust Hub on Jul 11, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill manages internal state and task briefs using local shell scripts (scripts/task-brief, scripts/review-package) that perform standard text processing and git operations. All operations are confined to the local repository and a dedicated scratch directory (.superpowers/sdd).
  • [SAFE]: Subagent interaction is governed by strict isolation principles. Implementation tasks are extracted into standalone brief files, and reviewers are restricted to analyzing generated diffs rather than crawling the codebase. This design minimizes the risk of context pollution and limits the impact of potentially malicious content in processed files.
  • [SAFE]: The workflow incorporates mandatory automated review cycles. A dedicated reviewer subagent must approve 'Spec Compliance' and 'Code Quality' before a task is marked complete, providing a robust internal control and validation mechanism.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 11, 2026, 07:04 PM