subagent-driven-development-extras

What problem does it solve?

This Skill eliminates the inefficiency of dispatching separate fix-and-re-review subagent cycles for every minor, non-blocking reviewer follow-up during multi-task subagent controller runs, which unnecessarily doubles subagent count for trivial changes. It also resolves common errors from missing pre-checks for non-main PR base branches and inaccurate severity model naming in security tool operator documentation that confuses operators during finding triage.

Core Features & Use Cases

• Two-tier review feedback routing: Automatically routes blocking review findings (request-changes verdicts) to a dedicated fix/re-review loop, while batching accepted non-blocking follow-ups into an explicit "COMMIT 1: review follow-ups" preamble in the next task's implementer prompt, cutting subagent overhead by roughly half for batches of minor feedback.
• Security documentation accuracy validation: Adds a quality reviewer check to confirm operator guides name the correct severity model for each security scanner (CVSS applies only to CVE-backed SCA dependency findings; SAST, IaC, and agent scanning use rule-based or tool-specific models) to prevent operator confusion.
• PR base branch pre-check: Includes a pre-verification step using git ls-remote --heads origin <base> to confirm non-main PR base branches exist on the remote origin before running gh pr create, avoiding the misleading "Base sha can't be blank" error from local-only base branches.
• Use Case: For teams running multi-task subagent-driven development workflows with quality review gates, this Skill eliminates redundant subagent dispatches, reduces documentation errors, and prevents common PR creation failures without adding manual overhead.

Quick Start

Use the subagent-driven-development-extras skill to route your quality reviewer's "approve with minor follow-ups" feedback for the current subagent task, batching all non-blocking items into the next dispatch's preamble commit.