Production-ready TypeScript starter template featuring multi-model routing (OpenAI, Anthropic, Ollama, LM Studio), execution history engine, zero runtime dependencies, and full Jest ESM test suite.
The built-in REPL with /help, /model, /clear, and /exit commands.
Everything you need to launch a custom AI coding CLI or embedded agent backend without reinventing infrastructure.
Switch seamlessly between OpenAI, Anthropic, Ollama, and LM Studio with priority fallback chains, request timeouts, and rate limiters.
Ultra-lean Node.js >=20 footprint. Built with standard Fetch API, native ESM modules, and zero heavy third-party framework bloat.
Includes comprehensive unit test coverage out of the box. Ships with instant TypeScript type checking and pre-configured dev tools.
import { InMemoryRouter } from './src/router.js';
// Initialize zero-dependency local model router
const router = new InMemoryRouter({
strategy: 'priority',
chain: [
{ name: 'ollama-local', endpoint: 'http://localhost:11434/v1', model: 'qwen2.5-coder' },
{ name: 'openai-cloud', endpoint: 'https://api.openai.com/v1', model: 'gpt-4o' }
]
});
// Execute model turn with history tracking
const response = await router.dispatch('Refactor src/router.ts for performance');
console.log(response.content);
Daedalus Lite's middleware architecture lets you intercept, enrich, and format AI interactions for any domain. Drop in a middleware function and you've got a specialized AI agent.
Scan git diffs for security vulnerabilities and generate structured PR checklists.
export const codeReviewer: Middleware = async (req, next) => {
req.prompt = `Review for security: ${req.prompt}`;
return next();
};
Inject knowledge base articles into prompts, log sentiment, and enforce response formatting.
export const supportRAG: Middleware = async (req, next) => {
const kb = await searchDocs(req.prompt);
req.prompt = `KB: ${kb}\nUser: ${req.prompt}`;
return next();
};
Prepend portfolio context and enforce regulatory disclaimers on every output.
export const financialGuard: Middleware = async (req, next) => {
const res = await next();
res.content += '\nNot financial advice.';
return res;
};
No subscriptions. No hidden fees. No per-seat licensing. Buy once, rebrand, and sell your own AI CLI product.
For indie developers
one-time purchase
For serious builders
one-time purchase
For teams & agencies
one-time purchase
Yes! Purchasing Daedalus Lite includes a commercial license allowing you to build, customize, and deploy proprietary or commercial software without royalties.
Daedalus Lite requires Node.js >= 20.0.0 for native fetch and ES module support.