| name | rdd-init |
| description | Initialize RDD framework in a new or existing project with TDD/BDD configuration |
RDD Init Skill
Initialize a new RDD (Roadmap Driven Development) project structure
Overview
The rdd-init skill sets up a complete RDD framework in a new or existing project directory. It creates the required directory structure, configuration files, documentation templates, and agent entry points needed for Roadmap Driven Development.
When to Use This Skill
Triggers:
- Starting a new project with RDD paradigm
- Converting an empty directory to an RDD project
- Setting up RDD structure after cloning a template
- After
/rdd-init command is invoked
- When project lacks
.rdd/ directory
Prerequisites:
- Empty or new project directory (for new projects)
- OR existing project directory (for migration - use rdd-migrate instead)
- Write permissions in target directory
- Git repository (recommended but not required)
Initialization Steps
Step 1: Pre-Initialization Check
Before initializing, verify:
[ ] Target directory exists and is accessible
[ ] No existing .rdd/ directory (if exists, skip or use rdd-migrate)
[ ] No conflicting RDD files (check for existing AGENTS.md, CLAUDE.md)
[ ] User has provided project name and description
If conflicts exist:
- Ask user if they want to overwrite existing files
- Suggest using rdd-migrate if project already has content
Step 2: Directory Creation
Create the following directory structure:
mkdir -p .rdd/cache
mkdir -p .rdd/scripts
mkdir -p .rdd/hooks
mkdir -p docs/framework
mkdir -p docs/stages
mkdir -p docs/handoff
mkdir -p .claude/skills
mkdir -p .claude/commands
Directory Purpose:
| Directory | Purpose |
|---|
.rdd/ | RDD configuration and runtime data |
.rdd/cache/ | Checkpoint state, temporary files |
.rdd/scripts/ | Hook notification scripts |
.rdd/hooks/ | Lifecycle hook scripts |
docs/framework/ | Governance and execution specifications |
docs/stages/ | Stage design documents |
docs/handoff/ | Agent handoff documents |
.claude/skills/ | Claude Code skill definitions |
.claude/commands/ | Claude Code command definitions |
Step 3: Configuration Files Setup
3.1 Create .rdd/config.yml
version: "1.0.0"
project:
name: "{{PROJECT_NAME}}"
description: "{{PROJECT_DESCRIPTION}}"
stage:
min_coverage: 20
max_failures: 3
tech_debt_threshold: 3
gates:
design_required: true
review_required: true
e2e_required: true
docs_required: true
hooks:
enabled: false
config_file: ".rdd/hooks.yml"
templates_file: ".rdd/templates.yml"
notifications:
quiet_hours:
enabled: true
3.2 Create .rdd/hooks.yml
channels:
wecom:
enabled: false
webhook_url: ""
email:
enabled: false
smtp_host: ""
smtp_port: 587
smtp_user: ""
smtp_pass: ""
from_address: ""
to_addresses: []
bark:
enabled: false
server_url: ""
device_key: ""
telegram:
enabled: false
bot_token: ""
chat_id: ""
webhook:
enabled: false
url: ""
method: "POST"
headers: {}
triggers:
roadmap_change:
enabled: true
channels: [wecom, email]
priority: "high"
[, , ]
[, ]
[]
[, ]
[, ]
[]
[]
3.3 Create .rdd/templates.yml
templates:
roadmap_change:
title: "Roadmap Updated"
block_message: "Roadmap change requires human review, Agent paused"
body: |
The project roadmap has been updated.
Project: {{project_name}}
Change Type: {{change_type}}
Changed By: {{changed_by}}
Timestamp: {{timestamp}}
Details:
{{change_details}}
failure_alert:
title: "Consecutive Failures Alert"
block_message: "{{failure_count}} consecutive failures, Agent paused"
body: |
CRITICAL: Multiple consecutive failures detected.
Project: {{project_name}}
Stage: {{stage_name}}
Failure Count: {{failure_count}}
Last Error: {{last_error}}
Timestamp: {{timestamp}}
hypothesis_invalid:
title: "Hypothesis Invalidated"
block_message: "Core hypothesis invalidated, Agent paused for decision"
body: |
A hypothesis has been invalidated during testing.
Project: {{project_name}}
{{}}
{{}}
{{}}
{{}}
{{}}
{{}}
{{}}
{{}}
{{}}
{{}}
{{}}
{{}}
{{}}
{{}}
{{}}
{{}}
{{}}
{{}}
{{}}
{{}}
{{}}
{{}}
{{}}
{{}}
3.4 Create .rdd/checkpoints.yml
checkpoints:
stage_complete_review:
enabled: true
trigger: "stage_complete"
requirements:
- "All tests passing"
- "Coverage meets minimum threshold"
- "No unresolved tech debt above threshold"
- "Documentation updated"
auto_approve_conditions:
coverage_above: 80
no_blockers: true
tech_debt_below: 1
adr_review:
enabled: true
trigger: "adr_created"
requirements:
- "ADR follows template format"
- "Decision rationale documented"
- "Alternatives considered"
- "Impact assessment included"
new_debt:
enabled: true
trigger: "tech_debt_added"
requirements:
- "Debt item documented"
- "Priority assigned"
- "Resolution plan proposed"
Step 4: Documentation Files Setup
4.1 Create docs/01-charter.md
# Project Charter
**Last Updated**: YYYY-MM-DD
## Vision
[One paragraph describing the project's ultimate goal]
## Goals
1. [Primary goal 1]
2. [Primary goal 2]
3. [Primary goal 3]
## Non-Goals
1. [Explicitly out of scope item 1]
2. [Explicitly out of scope item 2]
## Success Criteria
- [ ] [Measurable success criterion 1]
- [ ] [Measurable success criterion 2]
- [ ] [Measurable success criterion 3]
## Boundaries
**In Scope:**
- [What the project will do]
**Out of Scope:**
- [What the project will NOT do]
## Stakeholders
| Role | Responsibility |
|------|---------------|
| [Role 1] | [Responsibility] |
| [Role 2] | [Responsibility] |
## Timeline
- **Start Date**: YYYY-MM-DD
- **Target MVP**: YYYY-MM-DD
- **Target v1.0**: YYYY-MM-DD
4.2 Create docs/02-engineering-principles.md
# Engineering Principles
**Last Updated**: YYYY-MM-DD
## Core Principles
### 1. Roadmap Driven
- Humans define the roadmap; agents execute autonomously
- Roadmap changes require human approval
- Clear separation of strategy and execution
### 2. Stage as Delivery Unit
- Every stage is verifiable and rollback-capable
- Controlled scope with explicit boundaries
- Each stage validates a small set of hypotheses
### 3. Quality Gates
- Design before implementation
- Review before merge
- Tests before completion
- Docs before handoff
### 4. Explicit Knowledge
- All decisions recorded in ADRs
- All tech debt tracked in ledger
- All assumptions documented
### 5. Continuous Verification
- Multi-model cross-validation
- Real environment testing
- Clean environment verification
## Code Standards
- Follow language-specific style guides
- Write self-documenting code
- Maintain test coverage >= 20%
- Document public APIs
## Review Standards
- All code requires review
- Design documents require review
- Low-confidence findings need human verification
- Use triangulation for critical decisions
4.3 Create docs/03-stage-based-development.md
# Stage-Based Development
**Last Updated**: YYYY-MM-DD
## Overview
Stage-Based Development (SBD) breaks down project work into small, verifiable, and rollback-capable units called Stages.
## Stage Lifecycle
[GATE 0] Stage Startup
|
v
[GATE 1] Design Document
|
v
[GATE 2] Design Review
|
v
[GATE 3] Implementation & Testing
|
v
[GATE 4] Code Review
|
v
[GATE 5] Completion Check
## Stage Sizing Rules
- **Duration**: 1-3 days maximum
- **Scope**: Single focused feature or improvement
- **Hypotheses**: 1-3 core hypotheses per stage
- **Deliverables**: Testable, demonstrable output
## Completion Criteria
- [ ] All acceptance criteria met
- [ ] Unit tests pass (coverage >= 20%)
- [ ] E2E tests pass (at least 2 high-signal paths)
- [ ] Real environment verified
- [ ] Clean environment verified
- [ ] Design doc matches implementation
- [ ] Tech debt ledger updated
- [ ] ADR recorded
- [ ] fresh-agent-check passed
## Rollback Strategy
Each stage must define:
1. What version to rollback to
2. How to rollback
3. Data migration considerations (if any)
4.4 Create docs/08-autonomous-decisions.md
# Autonomous Decision Records (ADR)
**Last Updated**: YYYY-MM-DD
## Overview
This document records all significant autonomous decisions made during development. Each ADR captures the context, decision, and impact.
## ADR Index
| ID | Title | Stage | Date | Status |
|----|-------|-------|------|--------|
| - | (No decisions yet) | - | - | - |
---
## Decision Template
Use this template for new decisions:
```markdown
### Decision N: [Title]
**Background**: What circumstances led to this decision
**Decision**: What path was chosen
**Rationale**: Why this path was selected
**Impact on Subsequent Stages**: (Cannot be empty)
- [Specific impact on future work]
- [What this enables or constrains]
**Date**: YYYY-MM-DD
**Related Stage**: Stage N
**Alternatives Considered**:
1. [Alternative 1]: [Why not chosen]
2. [Alternative 2]: [Why not chosen]
#### 4.5 Create `docs/10-review-practices.md`
```markdown
# Review Practices
**Last Updated**: YYYY-MM-DD
## Multi-Model Review Methodology
### Triangulation Approach
1. **Main Model**: Primary development work
2. **Independent Review Model**: Fresh perspective
3. **Rule Checking**: Static analysis
### Verification Priority
When findings disagree:
1. **Authoritative sources** (documentation, specs)
2. **Code verification** (runtime behavior)
3. **Model inquiry** (ask a third model)
### False Positive Handling
- ~50% of findings are false positives
- Verify each finding independently
- Document false positives in review log
## Review Types
### Design Review (Gate 2)
Review design document before implementation:
- [ ] Goals are clear and specific
- [ ] Non-goals are explicitly stated
- [ ] Hypotheses are testable
- [ ] Acceptance criteria are measurable
- [ ] Rollback plan is defined
### Code Review (Gate 4)
Review implementation after E2E pass:
- [ ] Code follows design
- [ ] Tests cover critical paths
- [ ] No undocumented behavior
- [ ] Performance is acceptable
- [ ] Security concerns addressed
## Review Log Template
See `.claude/skills/rdd-templates/SKILL.md` for the full review log template.
4.6 Create docs/11-next-steps.md
# Next Steps
**Last Updated**: YYYY-MM-DD
## Current Status
**Active Stage**: Stage 0 (Initialization)
**Overall Progress**: 0%
## Immediate Actions
1. [ ] Define project vision and goals
2. [ ] Create initial roadmap
3. [ ] Set up development environment
4. [ ] Create Stage 0 design document
## Upcoming Stages
| Stage | Title | Priority | Status | Dependencies |
|-------|-------|----------|--------|--------------|
| 0 | Initialization | P0 | Planning | None |
| 1 | [TBD] | - | - | Stage 0 |
## Blockers
None currently.
## Notes
- This project was initialized with RDD framework
- Update this document as stages progress
4.7 Create docs/12-technical-debt.md
# Technical Debt Ledger
**Last Updated**: YYYY-MM-DD
## Overview
This document tracks all known technical debt. Each entry includes priority, source, impact, and resolution plan.
## Debt Summary
| Priority | Count | Blocking |
|----------|-------|----------|
| Architecture-level | 0 | 0 |
| Module-level | 0 | 0 |
| Local | 0 | 0 |
## Active Technical Debt
(No technical debt recorded yet)
---
## Tech Debt Template
Use this template for new debt:
```markdown
### TD-NN: Short Title
- **Priority**: [Blocking / Degraded Functionality / Technical Optimization] / [Architecture-level / Module-level / Local]
- **Source**: [Proactive prototype compromise / Review deferred / Autonomous decision compromise] (Stage N)
- **Original Description**: (Quote from original document)
- **Source File**: (File path and line number, if applicable)
- **Suggested Resolution Stage**: (Stage N or "Special iteration" or "As needed")
- **Impact**: [What this debt affects]
- **Resolution Cost Estimate**: [Low / Medium / High]
- **Created Date**: YYYY-MM-DD
- **Resolved Date**: (Empty until resolved)
#### Notes
[Additional context]
#### Resolution Plan
[When and how to address]
#### 4.8 Create `docs/stages/stage-roadmap.md`
```markdown
# Project Roadmap
**Last Updated**: YYYY-MM-DD
**Current Stage**: Stage 0
**Overall Progress**: 0%
---
## Vision
[Project vision - one paragraph describing the ultimate goal]
---
## Stage Overview
| Stage | Title | Status | Priority | Dependencies |
|-------|-------|--------|----------|--------------|
| 0 | Initialization | Planning | P0 | None |
---
## Stage Details
### Stage 0: Initialization
**Status**: Planning
**Goal**: Set up project foundation and infrastructure
**Prerequisites**: None
**Estimated Effort**: Small
---
## Current Focus
**Active Stage**: Stage 0
**Current Objective**: Initialize RDD framework and project structure
**Blockers**: None
**Next Milestone**: Complete Stage 0
---
## Roadmap History
| Date | Change | Reason |
|------|--------|--------|
| YYYY-MM-DD | Initial roadmap created | Project initialization |
---
## Priority Definitions
- **P0**: Must complete for MVP / Core functionality
- **P1**: Important for complete product experience
- **P2**: Nice to have, can be deferred
- **P3**: Future consideration
4.9 Create docs/stages/stage-template.md
# Stage Template
Use this template when creating new stage documents at `docs/stages/stage-N.md`.
---
# Stage N: [Title]
## Status
[ ] Planning / [ ] In Progress / [ ] Complete
## Goals
What this stage specifically solves (be precise and limited)
## Non-Goals
What this stage explicitly does NOT do (be explicit about scope boundaries)
## Core Hypotheses
- Hypothesis A: [Description of what you're trying to validate]
- Hypothesis B: [Description of what you're trying to validate]
## Acceptance Criteria
- [ ] Acceptance criterion A (must be testable)
- [ ] Acceptance criterion B (must be testable)
## Rollback Plan
Which version to fall back to if this stage fails
## Known Limitations
- [Limitation A - may become technical debt]
- [Limitation B - may become technical debt]
## Impact on Subsequent Stages
- [Impact A - what this stage enables or constrains]
- [Impact B - what this stage enables or constrains]
---
## Implementation Notes (Filled during implementation)
### Implementation Differences
[Document any differences from original design]
### Technical Decisions Made
[Document any technical decisions made during implementation]
### Testing Evidence
- Unit test coverage: X%
- E2E tests: [list test names]
- Real environment verification: [description]
- Clean environment verification: [description]
### Handoff Notes
[What the next agent needs to know]
Step 5: Agent Entry Points Setup
5.1 Create AGENTS.md
# Agent Entry Point
> This is the primary entry point for all AI agents working on this project.
## Mandatory Reading Order
1. **Start Here**: `docs/01-charter.md` - Project vision and goals
2. **Process**: `docs/03-stage-based-development.md` - How we work
3. **Principles**: `docs/02-engineering-principles.md` - Engineering standards
4. **Current State**: `docs/stages/stage-roadmap.md` - Where we are
5. **Next Actions**: `docs/11-next-steps.md` - What to do next
6. **Decisions**: `docs/08-autonomous-decisions.md` - Past decisions
7. **Debt**: `docs/12-technical-debt.md` - Known issues
## RDD Skills
- **rdd-core**: Core RDD concepts and workflow
- **rdd-templates**: Document templates for all artifacts
- **rdd-init**: Initialize new RDD projects
- **rdd-migrate**: Migrate existing projects to RDD
## Document Update Obligations
When completing work, you MUST update:
1. **Stage document** (`docs/stages/stage-N.md`) - Implementation differences
2. **Review log** (`docs/stages/stage-N-review-log.md`) - Review findings
3. **Decisions** (`docs/08-autonomous-decisions.md`) - New ADRs
4. () - New or resolved debt
() - Progress update
() - Change summary
Before completing a stage, verify:
[ ] All acceptance criteria met
[ ] Tests pass (unit + E2E)
[ ] Real environment verified
[ ] Clean environment verified
[ ] Design doc matches implementation
[ ] Tech debt ledger updated
[ ] ADR recorded with impact
[ ] fresh-agent-check passed
Use multi-model triangulation
Verify ~50% false positive rate
Document findings in review log
Escalate low-confidence findings
5.2 Create CLAUDE.md
# Claude Code Entry Point
> Quick reference for Claude Code when working with this RDD project.
## Quick Start
Initialize project (first time)
task bootstrap
Check project health
task doctor
Verify current stage
task stage:verify
Run gate checks
task stage:gate
## RDD Commands
| Command | Purpose |
|---------|---------|
| `/rdd-init` | Initialize new RDD project |
| `/rdd-migrate` | Migrate existing project |
| `/rdd-roadmap` | Manage roadmap |
| `/rdd-stage-auto` | Execute stage with gates |
| `/rdd-review-auto` | Automated review |
| `/rdd-knowledge` | Knowledge management |
| `/rdd-diagnosis` | Diagnose issues |
| `/rdd-fresh-check` | Verify fresh agent can take over |
## Key Files
| File | Purpose |
|------|---------|
| `AGENTS.md` | Agent entry point (start here) |
| `docs/01-charter.md` | Project vision |
| `docs/stages/stage-roadmap.md` | Current status |
| `docs/11-next-steps.md` | Immediate actions |
| `.rdd/config.yml` | RDD configuration |
## Gate Checklist
- **Gate 0**: Prerequisites verified
- **Gate 1**: Design doc created
- **Gate 2**: Design reviewed
- **Gate 3**: Implementation + tests
- **Gate 4**: Code reviewed
- **Gate 5**: Completion verified
## Tool Usage Rules
1. Read existing files before editing
2. Use Grep for searching code
3. Use Glob for finding files
4. Use Bash for git operations
5. Create commits with descriptive messages
5.3 Create CHANGELOG.md
# Changelog
All notable changes to this project will be documented in this file.
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/).
## [Unreleased]
### Added
- RDD framework initialized
- Project structure created
- Documentation templates added
## [0.1.0] - YYYY-MM-DD
### Added
- Initial project setup
- RDD framework integration
Step 6: Placeholder Files
Create .gitkeep files for empty directories:
touch .rdd/cache/.gitkeep
touch .rdd/scripts/.gitkeep
touch .rdd/hooks/.gitkeep
Step 7: Skills and Commands Setup
Create placeholder files for skills and commands directories:
touch .claude/skills/.gitkeep
touch .claude/commands/.gitkeep
Step 8: TDD/BDD Framework Setup (NEW)
CRITICAL: This step enforces 95%+ test coverage requirement from Stage 0.
8.1 Detect Project Language and Type
Run the test framework detection script:
./scripts/lib/test-framework-detect.sh detect
The script automatically detects:
- Language: Node.js, Python, Go, Rust, Java
- Project Type: frontend, backend-api, backend-service, fullstack
- Frameworks: React, Vue, Express, Django, FastAPI, etc.
8.2 Generate Test Configuration
./scripts/lib/test-framework-detect.sh config .rdd/test-config.yml
Creates .rdd/test-config.yml with:
- Recommended BDD framework
- Unit test runner
- Coverage tool
- E2E testing tool (if applicable)
- API testing tool (for backend projects)
8.3 Generate Example BDD Tests
./scripts/lib/test-framework-detect.sh feature tests/features
./scripts/lib/test-framework-detect.sh steps tests/steps
8.4 Framework-Specific Setup
Node.js Projects:
npm install --save-dev \
@cucumber/cucumber \
jest \
nyc \
supertest \
playwright
npm pkg set scripts.test="jest"
npm pkg set scripts.test:coverage="jest --coverage --coverageThreshold='{\"global\":{\"branches\":95,\"functions\":95,\"lines\":95,\"statements\":95}}'"
npm pkg set scripts.test:e2e="playwright test"
npm pkg set scripts.test:bdd="cucumber-js tests/features"
Python Projects:
pip install pytest pytest-bdd coverage pytest-cov requests
cat > pytest.ini << 'EOF'
[pytest]
testpaths = tests
python_files = test_*.py
python_classes = Test*
python_functions = test_*
addopts = --cov=src --cov-fail-under=95
EOF
Go Projects:
go install github.com/cucumber/godog/cmd/godog@latest
cat > godog.conf << 'EOF'
format: pretty
paths: tests/features
EOF
Rust Projects:
cat >> Cargo.toml << 'EOF'
[dev-dependencies]
cucumber = "0.20"
[[test]]
name = "cucumber"
harness = false
EOF
Java Projects:
<dependencies>
<dependency>
<groupId>io.cucumber</groupId>
<artifactId>cucumber-java</artifactId>
<version>LATEST</version>
<scope>test</scope>
</dependency>
<dependency>
<groupId>org.junit.jupiter</groupId>
<artifactId>junit-jupiter</artifactId>
<version>5.10.0</version>
<scope>test</scope>
</dependency>
</dependencies>
8.5 Configure Pre-commit Hooks
Create .pre-commit-config.yaml:
repos:
- repo: local
hooks:
- id: test-coverage
name: Test Coverage Check (95% minimum)
entry: bash -c 'npm run test:coverage || pytest --cov-fail-under=95'
language: system
pass_filenames: false
- id: lint
name: Lint Check
entry: npm run lint || ruff check .
language: system
pass_filenames: false
Install pre-commit:
pip install pre-commit
pre-commit install
npm install --save-dev husky lint-staged
npx husky install
8.6 Update RDD Configuration
Update .rdd/config.yml to enforce coverage:
gates:
design_required: true
review_required: true
e2e_required: true
docs_required: true
min_coverage: 95
coverage_fail_gate: true
bdd_required: true
8.7 Verify Test Setup
npm test
pytest
go test ./...
cargo test
mvn test
npm run test:coverage
pytest --cov
go test -cover
cargo tarpaulin
mvn test jacoco:report
Validation Checklist
After initialization, verify:
Directory Structure
[ ] .rdd/
[ ] .rdd/cache/
[ ] .rdd/scripts/
[ ] .rdd/hooks/
[ ] docs/
[ ] docs/framework/
[ ] docs/stages/
[ ] docs/handoff/
[ ] .claude/
[ ] .claude/skills/
[ ] .claude/commands/
Configuration Files
[ ] .rdd/config.yml
[ ] .rdd/hooks.yml
[ ] .rdd/templates.yml
[ ] .rdd/checkpoints.yml
Documentation Files
[ ] docs/01-charter.md
[ ] docs/02-engineering-principles.md
[ ] docs/03-stage-based-development.md
[ ] docs/08-autonomous-decisions.md
[ ] docs/10-review-practices.md
[ ] docs/11-next-steps.md
[ ] docs/12-technical-debt.md
[ ] docs/stages/stage-roadmap.md
[ ] docs/stages/stage-template.md
Entry Point Files
[ ] AGENTS.md
[ ] CLAUDE.md
[ ] CHANGELOG.md
Content Validation
[ ] .rdd/config.yml has project name placeholder
[ ] docs/01-charter.md has vision section
[ ] docs/stages/stage-roadmap.md has Stage 0
[ ] docs/11-next-steps.md has initialization tasks
[ ] AGENTS.md has mandatory reading order
TDD/BDD Validation (NEW)
[ ] .rdd/test-config.yml exists with correct framework
[ ] tests/features/ directory exists
[ ] tests/steps/ directory exists
[ ] Example feature file exists
[ ] Example step definitions exist
[ ] Test commands run successfully
[ ] Coverage threshold (95%) configured
[ ] Pre-commit hooks configured
Optional Git Setup
[ ] .gitignore includes .rdd/cache/
[ ] Initial commit created
Post-Initialization
After successful initialization:
-
Update project information:
- Edit
.rdd/config.yml with actual project name and description
- Edit
docs/01-charter.md with project vision
- Edit
docs/stages/stage-roadmap.md with initial stages
-
Configure notifications (optional):
- Edit
.rdd/hooks.yml to enable notification channels
- Set environment variables for credentials
-
Create first stage:
- Copy
docs/stages/stage-template.md to docs/stages/stage-0.md
- Fill in Stage 0 details for project initialization
-
Run health check:
- Execute
task doctor to verify setup
- Fix any reported issues
Troubleshooting
Directory Already Exists
If .rdd/ directory already exists:
- Check if project is already initialized
- Use
/rdd-migrate instead for existing projects
- Or manually remove
.rdd/ and reinitialize (WARNING: loses configuration)
Permission Denied
If write permissions are missing:
- Check directory ownership:
ls -la
- Fix permissions:
chmod -R u+w .
- Run from correct directory
YAML Parse Errors
If configuration files have syntax errors:
- Validate YAML:
python -c "import yaml; yaml.safe_load(open('.rdd/config.yml'))"
- Check for tab characters (use spaces)
- Verify indentation is consistent
Missing Dependencies
If Task commands fail:
- Install Task:
sh -c "$(curl --location https://taskfile.dev/install.sh)"
- Verify installation:
task --version
Reference
For more information:
.claude/skills/rdd-core/SKILL.md - Core RDD concepts
.claude/skills/rdd-templates/SKILL.md - Document templates
docs/plans/2026-03-06-rdd-framework.md - Implementation plan