agentsclimarketplace
Sign in

Rdd

Skill RubiYH/teamem/.agents/skills/rdd

Run review-driven development. Use when the user asks for RDD, review-driven development, or wants each issue handled by an executor-reviewer-debugger loop until reviewer approval before moving to the next issue.From its SKILL.md

Install
npx -y skills add RubiYH/teamem --skill rdd

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

2.1 KB, as published. Nobody here has run it

Review-Driven Development

Overview

Use this workflow to process issues one at a time with separated implementation, review, and fix roles. Keep each issue bounded, verify the result, and carry deferred out-of-scope findings to the final report.

Workflow

  1. Pick the next issue and define the expected done state.
  2. Summon a dedicated executor for only that issue.
  3. Summon a dedicated reviewer to inspect the executor's work.
  4. If the reviewer requests fixes, summon a debugger to address all in-scope fixes.
  5. Repeat review and debug until the reviewer approves.
  6. If reviewer-requested fixes are too far out of scope, defer them and report them after all issues finish.
  7. Discard the issue-specific agents and restart from step 1 with the next issue.
  8. Finish with changed files, verification evidence, deferred items, and remaining risks.

Role Rules

  • Keep each executor, reviewer, and debugger scoped to the current issue.
  • Only the main agent may summon RDD subagents, and each summon must happen from the main agent's context stream so every handoff is visible and auditable.
  • Executors, reviewers, debuggers, and other child agents must not summon additional subagents or delegate to sub-subagents.
  • Spawn issue-specific agents through oh-my-codex first when that surface is available.
  • If oh-my-codex subagents are unavailable, try Codex native subagents.
  • If neither oh-my-codex nor Codex native subagents are available, halt the RDD workflow and alert the user instead of emulating the roles in the main thread.
  • Do not let a reviewer implement fixes; route fixes through the debugger role.
  • Do not move to the next issue until the current reviewer approves or the remaining requests are explicitly deferred as out of scope.

Completion Report

Report:

  • Issues completed.
  • Files changed.
  • Tests or verification run.
  • Reviewer approval status.
  • Deferred out-of-scope requests.
  • Remaining risks.

Keep looking

Skills are one crate of 627,801. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.