{
  "id": "page:library-rpikit",
  "_kind": "Page",
  "_file": "wiki/library/rpikit.md",
  "_cluster": "wiki",
  "attributes": {
    "nodeKind": "Page",
    "title": "RPIKit Methodology (Library)",
    "displayName": "RPIKit Methodology (Library)",
    "slug": "library/rpikit",
    "articlePath": "wiki/library/rpikit.md",
    "article": "\n# RPIKit Methodology\n\n**Source**: [bostonaholic/rpikit](https://github.com/bostonaholic/rpikit) by Matthew Boston\n**Category**: Structured Software Development / Research-Plan-Implement Framework\n**License**: See upstream repository\n\n## Overview\n\nRPIKit implements the Research-Plan-Implement (RPI) framework for disciplined AI-assisted software development. It enforces structured progression through distinct phases with human approval checkpoints, stakes-based rigor scaling, and mandatory verification at every step.\n\n## Core Principles\n\n- **Iron Law**: \"Do NOT explore the codebase until the problem is understood\"\n- **Stakes-Based Rigor**: Planning intensity scales with change risk (low/medium/high)\n- **Test-First**: Every code-changing task specifies tests before implementation\n- **Verification-Before-Completion**: Never declare done without running verification\n- **Human Approval Gates**: Explicit human approval between major phases\n- **Soft-Gating Reviews**: Code and security reviews preserve user autonomy\n\n## Process Files\n\n| Process | File | Description | Task Count |\n|---------|------|-------------|------------|\n| Brainstorm | `rpikit-brainstorm.js` | Clarify vague requirements through exploration | 3 |\n| Research | `rpikit-research.js` | Systematic codebase exploration (Iron Law) | 5 |\n| Plan | `rpikit-plan.js` | Stakes-based implementation planning with test-first | 7 |\n| Implement | `rpikit-implement.js` | Disciplined plan execution with verification | 7 |\n| Review | `rpikit-review.js` | Combined code quality and security review | 4 |\n| Decision | `rpikit-decision.js` | Architecture Decision Records (ADRs) | 3 |\n\n## Skills Catalog\n\n| Skill | Directory | Description |\n|-------|-----------|-------------|\n| codebase-research | `skills/codebase-research/` | Systematic codebase exploration (Iron Law) |\n| plan-writing | `skills/plan-writing/` | Stakes-based implementation planning |\n| plan-implementation | `skills/plan-implementation/` | Disciplined plan execution with verification |\n| code-review | `skills/code-review/` | Structured code quality assessment |\n| security-review | `skills/security-review/` | Security vulnerability assessment |\n| brainstorming | `skills/brainstorming/` | Requirement clarification for vague goals |\n| decision-documentation | `skills/decision-documentation/` | Architecture Decision Records |\n| test-driven-development | `skills/test-driven-development/` | Test-first development practice |\n| systematic-debugging | `skills/systematic-debugging/` | Hypothesis-driven defect investigation |\n| verification | `skills/verification/` | Verification-before-completion discipline |\n| finishing-work | `skills/finishing-work/` | Final completion and summary protocol |\n\n## Agents Catalog\n\n| Agent | Directory | Role |\n|-------|-----------|------|\n| file-finder | `agents/file-finder/` | File Discovery Specialist |\n| web-researcher | `agents/web-researcher/` | External Context Gatherer |\n| code-reviewer | `agents/code-reviewer/` | Code Quality Assessor |\n| security-reviewer | `agents/security-reviewer/` | Security Vulnerability Assessor |\n| test-runner | `agents/test-runner/` | Test Execution Specialist |\n| verifier | `agents/verifier/` | Implementation Verifier |\n| debugger | `agents/debugger/` | Defect Investigator |\n\n## Workflow Lifecycle\n\n```\n[Optional] Brainstorm (\"what to build\") -> Research (Iron Law) -> Plan (stakes-based) -> Implement (verify each step) -> Review (code + security) -> Done\n                                                                                                     ^                              |\n                                                                                                     +--- Decision (ADR at any point) ---+\n```\n\nCross-cutting concerns applied throughout:\n- `verification` - Every step verified before proceeding\n- `test-driven-development` - Tests specified before implementation\n- `systematic-debugging` - Failures investigated with hypothesis-driven methodology\n- `finishing-work` - Completion confirmed against success criteria\n\n## Stakes Classification\n\n| Level | Profile | Planning Approach |\n|-------|---------|-------------------|\n| Low | Isolated, reversible, minimal impact | Brief plan, inline if needed |\n| Medium | Multiple files, moderate scope, testable | Standard RPI cycle |\n| High | Architectural, difficult rollback, security-sensitive | Comprehensive research and detailed planning required |\n\n## Output Artifacts\n\n- `docs/plans/YYYY-MM-DD-<topic>-research.md` - Research findings\n- `docs/plans/YYYY-MM-DD-<topic>-plan.md` - Implementation plans\n- `docs/decisions/NNNN-<title>.md` - Architecture Decision Records\n\n## Philosophy\n\n- **Understand before acting** - Research before planning\n- **Plan before coding** - Plans before implementation\n- **Implement with discipline** - Step-by-step verification\n- **Scale rigor to risk** - Stakes-based planning depth\n- **Preserve autonomy** - Reviews are soft gates, humans decide\n",
    "documents": [
      "specialization:rpikit"
    ]
  },
  "outgoingEdges": [
    {
      "from": "page:library-rpikit",
      "to": "specialization:rpikit",
      "kind": "documents"
    }
  ],
  "incomingEdges": [
    {
      "from": "page:index",
      "to": "page:library-rpikit",
      "kind": "contains_page"
    }
  ]
}