Quick start
Prerequisites
- Node.js 18+
- pnpm
- An API key for your chosen model (e.g.
OPENAI_API_KEYfor OpenAI)
Use as a package
pnpm add @lablnet/hyperagents
Import agents, loop, domains, and utilities from @lablnet/hyperagents. See API reference.
Run from this repository
git clone https://github.com/Framework-Island/hyperagents.git
cd hyperagents
pnpm install
cp .env.example .env # set OPENAI_API_KEY (and others if needed)
Scoring demo (recommended first run)
pnpm demo:scoring
Watch average score improve as the MetaAgent fixes prompt logic (e.g. math equivalence vs strict string match).
Other examples
| Command | What it shows |
|---|---|
pnpm example:bash | Bash command generation |
pnpm demo:calculator | Tool code fixed by evolution |
pnpm example:paper-review | Accept/reject style task |
npx tsx examples/bash/run.ts evolve | Full loop on bash domain |
pnpm demo:git-evolution | Git-based patches, full loop |
LLM providers (library)
import { createLLM } from "@lablnet/hyperagents";
createLLM({ model: "openai/gpt-4o" });
createLLM({ model: "anthropic/claude-sonnet-4-5-20250929" });
createLLM({ model: "gemini/gemini-2.5-pro" });
createLLM({ model: "ollama/llama3" });
Docker (optional)
From the repo root, build and run as in the main README: mount volumes for keys and outputs as needed.
Documentation site
cd docs_site
pnpm install
pnpm docs:dev
Then open the URL VuePress prints (usually http://localhost:8080).
Further reading
- Concepts — archive, harness, parent selection, prompt files.
- Limitations — cost, frozen model, framework boundaries.
