amanpal3/Dr-SKILLs / .agent/plugins/unified-build-pipeline/skills/unified-build-pipeline
unified-build-pipeline
VerifiedUnified autonomous development pipeline combining GSD (spec & context engineering), Ralph Loop (deterministic story execution & browser testing), and CodeRabbit (AI code review & quality gates). Use when orchestrating end-to-end development, running autonomous build loops with automated code review, or planning full-stack features.
Unified Autonomous Build Pipeline: GSD + Ralph Loop + CodeRabbit
The 5-Stage Unified Lifecycle
[ GSD Spec & Architecture ]
│
▼
[ Ralph Deterministic PRD ]
│
▼
[ Ralph Build Loop Execution ]
│
▼
[ CodeRabbit Automated Review ]
│
▼
[ GSD State & Git Commit ] ───► (Next Story)
Stage 1: GSD Discovery & Context Engineering
- Initialize Project / Feature:
- Run
/new-projector/mapto generate.gsd/SPEC.md,.gsd/ROADMAP.md, and.gsd/ARCHITECTURE.md.
- Run
- Context Boundaries:
- Enforce the Planning Lock: No implementation before
SPEC.mdis approved. - Set context limits: Keep turns under 50-70% context budget (
gsd-context-health-monitor).
- Enforce the Planning Lock: No implementation before
Stage 2: Ralph Deterministic PRD Generation
- Convert GSD Spec into Ralph PRD:
- Load
ralph-prdto translate the phase requirements into.agents/tasks/prd.json(or.ralph/prd.json).
- Load
- Story Rules:
- Atomic stories (
US-001,US-002, ...), each completable in a single iteration. - Explicit
qualityGates(e.g.npm test,npm run lint,npm run build). - Verifiable acceptance criteria including 1 positive example and 1 negative edge case.
- Atomic stories (
Stage 3: Ralph Autonomous Build Loop
For each story iteration:
- Pick Story: Select first open story where all
dependsOnare passed. - Load Guardrails: Check
.ralph/guardrails.mdand.ralph/errors.log. - Atomic Implementation: Code only the selected story scope.
- UI Verification:
- For frontend tasks, launch dev server and verify in browser using
dev-browser.
- For frontend tasks, launch dev server and verify in browser using
Stage 4: CodeRabbit AI Code Review Gate
Before marking any story complete or committing:
- CodeRabbit Static & Quality Audit:
- Security Check: Injection, auth bypass, secret leaks, unvalidated inputs.
- Performance Check: Unnecessary re-renders, memory leaks, unindexed queries.
- Regression Check: Backward compatibility with existing modules.
- Idiomatic Style: YAGNI / minimal code enforcement (via
ponytail).
- Auto-Fix: Remediate any high or medium severity findings immediately.
- Run Quality Gates: Execute
npm test,npm run lint,npm run typecheck.
Stage 5: Atomic Commit & State Persistence
- Git Commit:
git add -A git commit -m "feat(US-XXX): <story-title> [reviewed by CodeRabbit]" - Update Logs & State:
- Append execution summary to
.ralph/progress.md. - Update
.gsd/STATE.mdwith persistent memory and decisions. - Mark story
status: "passed"inprd.json.
- Append execution summary to
- Loop: Output
<promise>COMPLETE</promise>and proceed to the next story.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the work | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixes | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation code | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |