Introduction

Fuze is runtime safety middleware for AI agents. It wraps any existing framework — LangGraph, CrewAI, Google ADK, raw OpenAI/Anthropic SDK — and provides loop detection, budget enforcement, side-effect tracking, and EU AI Act-compliant audit trails.

What Fuze does

Fuze is not a framework. It's a middleware layer that wraps your existing agent tool calls:

• Loop detection: 5-layer detection catches ping-pong patterns, semantic stalls, and repetitive tool calls
• Budget enforcement: hard token, cost, and time ceilings per run with kill switches
• Side-effect tracking: knows which tool calls changed the real world, with compensation functions for rollback
• Audit trails: full replayable trace of every agent decision, Art. 12 compliant
• Smart recovery: retry with modified prompt, rollback to checkpoint, fork to alternate path, or escalate to human
• MCP Proxy: wrap any MCP server with zero code changes

How it works

Wrap any tool function with guard(), or group multiple steps into a run with createRun(). Fuze handles the rest.

import { guard, createRun } from 'fuze-ai'

const search = guard(async (query: string) => {
  return await vectorDb.search(query)
})

// Or group multiple tool calls into a tracked run
const run = createRun({ maxCost: 5.00, maxIterations: 50 })
const results = await run.call(search, 'quarterly revenue')

That's it. Every guarded function and every run has loop detection, budget tracking, and audit logging built in.

Deployment modes

Next steps

• Quickstart — get running in 30 seconds
• guard() API — full API reference
• Configuration — fuze.toml reference
• Examples — end-to-end usage examples