v1.0.0

2026-07-21 18:33:22 -07:00 · 2025-12-04 17:34:22 -08:00
parent 4086c2e592
commit 15911e970a
38 changed files with 12292 additions and 0 deletions
@@ -0,0 +1,83 @@
+# CLAUDE.md
+
+This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
+
+## What This Repository Is
+
+Plan2Code is a structured 4-step workflow methodology for AI-assisted software development. It contains prompt templates, not executable code. The workflow emphasizes thorough planning before implementation.
+
+## Development Commands
+
+```bash
+# Sync source prompts to all platform-specific directories
+node sync-prompts.js
+
+# Preview changes without writing
+node sync-prompts.js --dry-run
+```
+
+## Repository Architecture
+
+**Source prompts** (root directory) are the canonical versions:
+- `plan2code-1--plan.md` - Planning mode prompt
+- `plan2code-2--document.md` - Documentation mode prompt
+- `plan2code-3--implement.md` - Implementation mode prompt
+- `plan2code-4--finalize.md` - Finalization mode prompt
+
+**sync-prompts.js** copies these to platform-specific directories with appropriate YAML headers:
+- `.claude/commands/` - Claude Code CLI (no headers)
+- `.cursor/rules/` - Cursor AI (`.mdc` extension, `alwaysApply: false`)
+- `.github/prompts/` - VS Code Copilot (`mode: agent`)
+- `.github/agents/` - GitHub Copilot CLI
+- `.continue/prompts/` - Continue extension
+- `.windsurf/workflows/` - Windsurf IDE
+- `.agent/workflows/` - Google Antigravity
+
+**Always edit source prompts in root, then run `node sync-prompts.js` to propagate changes.**
+
+## Workflow Steps
+
+1. **Plan** (`plan2code-1--plan.md`) - Requirements analysis and architecture design as a senior architect
+2. **Document** (`plan2code-2--document.md`) - Transform planning output into structured implementation docs
+3. **Implement** (`plan2code-3--implement.md`) - Execute implementation phase by phase
+4. **Finalize** (`plan2code-4--finalize.md`) - Validate, summarize, and archive
+
+## Key Behavioral Rules
+
+- Start a **new conversation** before each step (and for each implementation phase)
+- Complete only **one planning/implementation phase at a time**, then stop
+- Must reach **90% confidence** before finalizing plans
+- **Tech stack decisions require explicit user approval**
+- Keep **checkboxes updated** in spec files for progress tracking
+- Follow specifications **exactly as documented**
+- Do NOT run tests unless explicitly included in phase tasks
+
+## Response Prefixes
+
+Each mode has a required prefix:
+
+- Planning: `🤔 [CURRENT PLANNING PHASE]`
+- Documentation: `📝 [CURRENT DOCUMENTATION PHASE]`
+- Implementation: `⚡ [CURRENT IMPLEMENTATION PHASE]`
+- Finalization: `🧹 [FINALIZATION STEP]`
+
+## Output Structure
+
+Specs are stored in `specs/` folder:
+
+```
+specs/
+├── PLAN-DRAFT-<timestamp>.md    # From Step 1
+├── <feature-name>/
+│   ├── overview.md              # Phase checklist
+│   └── Phase X.md               # Detailed tasks per phase
+└── completed/                   # Archived after finalization
+```
+
+## Using as Slash Commands
+
+In Claude Code, invoke with:
+- `/plan2code-1--plan` - Start planning
+- `/plan2code-2--document` - Create implementation docs
+- `/plan2code-3--implement` - Execute implementation
+- `/plan2code-4--finalize` - Validate and archive