Token-efficient agentic TypeScript runtime

Docs · API Reference · Examples

---

Six primitives. One agent loop. No magic. Flint gives you well-typed building blocks for AI agents in TypeScript and stays out of the way — plain async functions over the provider API, one runtime dependency, errors returned as values instead of thrown.

Install

npm install flint @flint/adapter-anthropic

ESM only, Node 20+, strict: true. The Anthropic adapter reads ANTHROPIC_API_KEY from the environment. See the setup guide for tsconfig and .env details.

Quick start

import { call, tool, agent } from 'flint';
import { budget } from 'flint/budget';
import { anthropicAdapter } from '@flint/adapter-anthropic';
import * as v from 'valibot'; // any Standard Schema library works

const adapter = anthropicAdapter({ apiKey: process.env.ANTHROPIC_API_KEY! });

// One-shot call — errors come back as values, nothing throws
const res = await call({
  adapter,
  model: 'claude-opus-4-7',
  messages: [{ role: 'user', content: 'What is the capital of France?' }],
});
if (res.ok) console.log(res.value.message.content); // "Paris"

// A tool
const add = tool({
  name: 'add',
  description: 'Add two numbers',
  input: v.object({ a: v.number(), b: v.number() }),
  handler: ({ a, b }) => a + b,
});

// Agent loop with enforced step / token / dollar caps
const out = await agent({
  adapter,
  model: 'claude-opus-4-7',
  messages: [{ role: 'user', content: 'What is 123 + 456?' }],
  tools: [add],
  budget: budget({ maxSteps: 5, maxDollars: 0.10 }),
});
if (out.ok) console.log(out.value.message.content); // "579"

What you get

Core (flint) — 1 runtime dependency (@standard-schema/spec)

• 6 primitives: call, stream, validate, tool, execute, count
• agent() loop with step / token / dollar budget caps
• compress transforms + pipeline(): dedup, windowLast, windowFirst, truncateToolResults, summarize, orderForCache
• recipes: react, retryValidate, reflect, summarize
• RAG (chunk, store, retrieve), conversation memory with async summarization
• safety: injection detection, redaction, permissions, approval gates, boundary wrapping

Packages

Package	Description
flint	Core primitives, agent loop, compress, memory, RAG, safety, recipes
@flint/adapter-anthropic	Anthropic Messages API — prompt-cache aware
@flint/adapter-openai-compat	Any OpenAI-compatible endpoint (OpenAI, Groq, Ollama, DeepSeek, Together)
@flint/graph	State-machine agent workflows with memory checkpointing
@flint/landlord	Multi-agent orchestration — dynamic workflow runtime and auto-decompose orchestrate()

Adapters and graph each ship with zero runtime dependencies, on Web API primitives only (fetch, ReadableStream, TextDecoder).

Why Flint

• One dependency — @standard-schema/spec only. No transitive framework sprawl.
• No classes, no chains — plain functions that compose naturally.
• Standard Schema — bring Zod, Valibot, ArkType, or any compatible library.
• Budget-aware — every agent loop enforces step, token, and dollar limits.
• Streaming first — AsyncIterable<StreamChunk> throughout.
• Safety in core — injection detection, redaction, and approval gates, not an afterthought.
• Results, not exceptions — Promise<Result<T>> everywhere; no try/catch at the call site.

Coming from LangChain? See the Flint vs LangChain guide for a full side-by-side.

Documentation

Full docs at dizzymii.github.io/Flint — primitives, features, adapters, examples, landlord, testing, and FAQ.

Contributing

See CONTRIBUTING.md.

License

MIT — see LICENSE.