flytohub/flyto-core

Overview

An open-source execution engine for AI agents. It provides 412 modules, is MCP-native, and includes triggers, a queue, versioning, and metering. The engine orchestrates agent workflows with these built-in features.

Best for

Best for
Developers building custom AI agent systems that need orchestration and metering

Use cases

• Building agent workflows with trigger-based execution
• Managing agent runs with queue and versioning
• Tracking usage and costs with built-in metering

How to use

Install

pip install flyto-core[browser] && playwright install chromium

Notes

An open-source execution engine for AI agents. It provides 412 modules, is MCP-native, and includes triggers, a queue, versioning, and metering. The engine orchestrates agent workflows with these built-in features.

272 stars on GitHub. Last updated 2026-05-31. Licensed Apache-2.0.

Use cases

• Building agent workflows with trigger-based execution
• Managing agent runs with queue and versioning
• Tracking usage and costs with built-in metering

Pros

• Open-source with full code access
• MCP-native for interoperability with other tools
• Includes versioning and metering out of the box

Cons

• Small community with 272 GitHub stars
• Python-only, limiting language choice
• May lack extensive documentation or support

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.