cli-five
v0.2.13
Published
Code Like I'm Five — scaffold a 5-agent VS Code Copilot team into any repo.
Downloads
2,650
Maintainers
Readme
cli-five
Code Like I'm Five — scaffold a 5-agent VS Code Copilot team into any repo.
Two ways to install
Full setup (recommended for teams)
npx cli-five initInterviews you, scaffolds agents + memory files + stack-specific instructions, discovers skills from awesome-copilot and skills.sh, and tells you what to do next.
Quick plugin install (personal use)
copilot plugin install idusortus/cli-fiveInstalls the 5 agents to your Copilot profile. No project config, no interview — just the agents with the same autonomous contracts shipped by the scaffolded templates.
What you get
your-repo/
├── .github/
│ ├── agents/ # 5 agents — Orchestrator delegates autonomously
│ │ ├── orchestrator.agent.md # Routes tasks: Planner → Coder/Designer → Reviewer
│ │ ├── planner.agent.md # Researches codebase, produces implementation plans
│ │ ├── coder.agent.md # Writes production code, runs tests
│ │ ├── designer.agent.md # UI/UX design, layouts, theming
│ │ └── reviewer.agent.md # Code review, convention compliance
│ ├── copilot-instructions.md # Persona + project mandates from interview
│ ├── instructions/ # Stack-specific coding guidelines
│ └── skills/ # Installed skills from awesome-copilot + skills.sh
├── AGENTS.md # Tool-agnostic project context (agents.md standard)
├── PROJECT.md # Long-form vision (rarely changes)
├── STATE.md # Cross-session status (changes constantly)
├── decisions.md # Architectural decision log
├── agent-diary.md # Append-only session log
└── histories/ # Per-agent accumulated learningsHow it works
Select the Orchestrator agent in Copilot Chat and describe what you want. The Orchestrator autonomously calls Planner → Coder/Designer → Reviewer without any manual handoff clicks.
User: implement from plan.md
Orchestrator → [calls Planner] → [calls Coder] → [calls Reviewer] → doneNo buttons. No "click here to continue". Just results.
The plugin agents and scaffolded .github/agents/*.agent.md templates are intentionally kept in sync. The repository test suite now checks that parity so the quick-install path does not quietly degrade.
Commands
npx cli-five init # interview + scaffold + skill discovery
npx cli-five doctor # validate an existing setup
npx cli-five list-stacks # show detectable tech stacks
npx cli-five helpFlags
| Flag | Effect |
|---|---|
| --yes, -y | Accept interview defaults (overwrite gate still active) |
| --force | Overwrite without confirmation (use with --yes) |
| --dry-run | Print actions, write nothing |
| --no-skills | Skip skills.sh discovery step |
| --cost-mode <m> | Set cost mode: premium, cheap, or mixed |
| --cwd <path> | Run against a different directory |
What init does
- Detect — fingerprints stack (Node/TS, Python, .NET, Kotlin, Rust, Go, etc.). Brownfield-aware.
- git init — if needed. Asks first.
- Overwrite gate — double-confirms ("Proceed?" then "R U Sure?"). Only
--force --yesbypasses. - Interview — name, one-liner, stack, frameworks, goals, constraints, cost mode, persona toggle.
- Scaffold — writes 18 files. Substitutes answers into templates. Swaps
model:per cost mode. - Skill discovery — the hero feature. Multi-source discovery from awesome-copilot (30k+ ★ community marketplace) and skills.sh. Color-coded recommendations, source attribution, multiselect install, and post-install breadcrumbs with copy-paste commands for the awesome-copilot suggestion skill and MCP server.
- Custom instructions — generates stack-specific
.instructions.mdfiles for detected languages. - Next steps — tells you to open Copilot Chat, select Orchestrator, and start building. No button-clicking required.
Cost modes
| Agent | premium (1x–3x) | cheap (0x) | mixed |
|---|---|---|---|
| Orchestrator | Claude Sonnet 4.6 | GPT-4.1 | GPT-4.1 |
| Planner | Claude Opus 4.6 | GPT-4o | GPT-4o |
| Coder | GPT-5.3-Codex | GPT-4.1 | GPT-5.3-Codex |
| Designer | Claude Opus 4.6 | GPT-4o | GPT-4o |
| Reviewer | Claude Opus 4.6 | GPT-5 mini | Claude Opus 4.6 |
Override from CLI: npx cli-five init --cost-mode cheap
Change anytime by editing the model: line in .github/agents/*.agent.md.
Required VS Code settings
{
"chat.subagents.allowInvocationsFromSubagents": true
}Skill discovery
cli-five searches two sources for skills matching your detected stack:
| Source | What it has | Stars | |---|---|---| | awesome-copilot | Skills, instructions, agents, plugins from the GitHub community | 30k+ | | skills.sh | Curated skill repos (Vercel, Anthropic, Microsoft, etc.) | — |
Recommendations show with source attribution. Only cli-five core skills are pre-selected; suggested skills from skills.sh and awesome-copilot require manual selection (press Space) before install. After installation, you get breadcrumbs for ongoing discovery:
Take time to read each installed skill so you understand what it does and can catch overlap or conflicts before they affect your workflow.
- Suggest skill —
copilot plugin install awesome-copilot@suggest(AI-driven repo analysis) - MCP server —
awesome-copilot-mcpfor programmatic search from any agent - CLI browser —
npx skills findfor interactive search
What this is not
- Not a runtime. Once scaffolded, your repo doesn't depend on
cli-five. You can uninstall the package and the agents still work. - Not a workflow engine. No
/plan-phase, no/ship, no 50 slash commands. Use/agent-customizationin Copilot Chat to extend. - Not Ralph. No issue-polling daemons, no autonomous loops, no crypto tokens.
Why the name
ELI5 → CLI5. Code Like I'm Five. Five agents. Get it? Yeah, it's a stretch. But the npm name was available.
Influences
- bradygaster/squad — see docs/squad-review.md
- burkeholland/ultralight — see docs/ultralight-review.md
- gsd-build/get-shit-done — see docs/gsd-review.md
Notes
- The plugin install path and the scaffolded project path are both supported and now validated against each other.
npm testincludes agent integrity checks in addition to the existing skill-step tests.
Local development
git clone https://github.com/idusortus/cli-five
cd cli-five
npm install
node bin/cli-five.mjs init --cwd /tmp/test-targetLicense
MIT.
