Skip to content

⚡ 1. Quickstart & Live Walkthrough in L1DDER ​

⬅️ Back to Main Map (README.md) | ➡️ Next: Flows in Detail (02-FLOWS-IN-DETAIL.md)


Step 1. Install L1DDER in Your Project (30 Seconds) ​

You don't even have to open a terminal — just paste this single line into your AI agent chat (Cursor, Claude Code, Codex, Windsurf, Copilot):

text
Install the L1DDER framework in this project by running `npx l1dder@latest init --lang en --yes`, then immediately start onboarding from `l1dder/skills/l1-onboard/SKILL.md`

Or run it in your project terminal:

bash
# Interactive setup (asks for language en/ru and your AI agents)
npx l1dder init

# Or non-interactive one-liner (defaults to Cursor, Claude Code & Codex):
npx l1dder init --lang en --yes

Then open your agent chat and run:

text
/l1-onboard

The agent walks you through six stages: idea, vision, architecture, data model, stack, Idea Bank and roadmap. At each stage it asks up to 7 questions with ready recommendations and asks whether to move on or discuss more. It proposes the stack only after the architecture and records it in l1dder/rules/30-tech-stack.mdc. All slash commands (/l1-onboard, /l1-idea, /l1-feature, /l1-planning, /l1-task, /l1-hotfix, /l1-chore, /l1-fast, /l1-evo) are ready in your chat.


Step 2. Live Day-to-Day Walkthrough ​

Suppose we are building a knowledge base web app and you get an idea: "I want instant article search and tag filtering synced to the URL".

💬 Scene 1: Capturing the Idea (/l1-idea) ​

  • You: /l1-idea I want instant search by article title and tag filtering synced to the URL
  • Agent: Checks docs/ideas.md, asks one decision-tree question (❓ Q1 — Multi-tag combination ➡️ Recommendation: AND logic with URL query ?tags=api,auth), and shows draft card IDEA-002.
  • You: ok → Saved in docs/ideas.md as 📋 unrealized.

💬 Scene 2: Feature Specification (/l1-feature) ​

  • You: /l1-feature IDEA-002
  • Agent: Proposes FEAT-002 (requirements.md, design.md with Deep Module CatalogFilter and Pre-agreed Test Seams, and tasks.md with FEAT-002.01, FEAT-002.02, and mandatory FEAT-002.99 Doctor Checklist).
  • You: ok → Spec created in docs/specs/FEAT-002-catalog-search/, added to docs/backlog.md (📋 in-backlog), and IDEA-002 becomes 🚧 in-progress.

💬 Scene 3: Scheduling on the Roadmap (/l1-planning) ​

  • You: /l1-planning
  • Agent: Places SPRINT-02 (FEAT-002 ← IDEA-002) at 1 (CURRENT) in docs/roadmap.md and assigns TASK-02.01, TASK-02.02, TASK-02.99.
  • You: ok → docs/roadmap.md updated and docs/sprints/SPRINT-02-catalog-search/README.md created.

💬 Scene 4: Executing Tasks & Sprint Doctor (/l1-task) ​

  • You: /l1-task
  • Agent: Reads docs/roadmap.md, writes plan.md for TASK-02.01, stops at ⛔ STOP #1 (ok), implements code & tests, runs self-check, stops at ⛔ STOP #2 (ok), writes summary.md, and marks TASK-02.01 as ✅ realized.
  • When regular tasks finish, /l1-task runs TASK-02.99 (Doctor Checklist): runs make ci, 2-axis review, E2E checks, syncs living docs (docs/concept/, l1dder/rules/), and gives you a Human UAT checklist. After your "ok", SPRINT-02, FEAT-002, and IDEA-002 cascade to ✅ realized!