<!-- mobian-agent-page publisher="dailydev" canonical="https://daily.dev/posts/the-opensource-managed-agents-hh62g6j2v" -->

---
title: The OpenSource Managed Agents | daily.dev
description: True Forge is an open-source, MIT-licensed, self-hosted agent execution harness from True Foundry that replicates the functionality of hosted 'managed agents'...
canonical: https://daily.dev/posts/the-opensource-managed-agents-hh62g6j2v
twitter:card: summary_large_image
twitter:site: @dailydotdev
og:type: website
og:site_name: daily.dev
og:title: The OpenSource Managed Agents | daily.dev
og:description: True Forge is an open-source, MIT-licensed, self-hosted agent execution harness from True Foundry that replicates the functionality of hosted 'managed agents'...
og:url: https://daily.dev/posts/the-opensource-managed-agents-hh62g6j2v
og:image: https://api.daily.dev/og/posts/hh62g6J2V.png
og:image:alt: The OpenSource Managed Agents
og:image:width: 1200
og:image:height: 630
og:locale: en
---

> ## Documentation Index
> Fetch the complete documentation index at: https://daily.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# The OpenSource Managed Agents

**[Sam Witteveen](https://daily.dev/sources/samwitteveenai)** · 20 min read · 0 upvotes · 0 comments

## Summary

True Forge is an open-source, MIT-licensed, self-hosted agent execution harness from True Foundry that replicates the functionality of hosted 'managed agents' (like Claude or Gemini managed agents) but lets you run any model, on your own hardware or private cloud, without vendor lock-in or a runtime bill. The video explains three categories of agent harnesses (coding harnesses, assemble-it-yourself frameworks like LangGraph, and hosted runtime/managed agents), then argues that the harness itself, not the model, is usually the biggest driver of agent costs. It cites a True Foundry benchmark on 14 DevRev enterprise-bench tasks showing True Forge using 3.8M tokens versus Claude managed agents' 10M tokens on the same model (Opus), a 30% cost cut with equal accuracy, and up to 75% cost reduction when switching to GLM 5.2. The video walks through installing True Forge locally via npx, configuring models (including DeepSeek V4.1 Flash), connectors (Tavily, DeepWiki, custom MCP servers), skills, and building/testing an AI news agent, plus calling it via a Python SDK.

## Full article

daily.dev links to this article rather than hosting it. Read it at the original source: <https://www.youtube.com/watch?v=3AdW345Jskk>

## Questions this post answers

### How much cheaper is True Forge compared to Claude managed agents for the same tasks?

True Forge cut costs by about 30% compared to Claude managed agents while maintaining the same accuracy, using the same Opus model. In a benchmark of 14 tasks from DevRev's enterprise bench, Claude managed agents used about 10 million tokens per run on average versus only 3.8 million tokens with True Forge. Switching to GLM 5.2 on True Forge brought total cost down 75% versus the original Claude managed agents setup.

_Developers weighing managed versus self-hosted agent harnesses can track cost benchmarks like this on daily.dev._

### What is True Forge and how does it differ from Claude or Gemini managed agents?

True Forge is an open-source, MIT-licensed agent execution harness from True Foundry that provides the same capabilities as hosted managed agent platforms (session handling, tool calls, sandboxes, approvals, streaming, a UI) but runs fully locally or in a private cloud with any model of choice, unlike Claude or Gemini managed agents which lock you into their own models and hosted infrastructure.

_Teams comparing self-hosted and vendor-managed agent runtimes can follow tools like this on daily.dev._

### How do I get started running True Forge locally?

True Forge is a TypeScript repo installable with a single npx command, which sets up a local SQLite database so all data stays on your machine, with hosted modes available using Postgres, Redis, and Docker. After installation, models, connectors (like Tavily or custom MCP servers), skills, and sandbox providers are configured through a settings UI, and agents can be tested via a built-in chat interface or called through an HTTP API and TypeScript SDK.

_Developers setting up self-hosted agent tooling can find practical walkthroughs like this on daily.dev._

## Similar posts on daily.dev

- [Preloading Knowledge Into a Model Instead of Retrieving It](https://daily.dev/posts/preloading-knowledge-into-a-model-instead-of-retrieving-it-nzwsexx0x) · Daily Dose of Data Science \| Avi Chawla \| Substack · 1 upvotes · 0 comments

---

Tags: [#open-source](https://daily.dev/tags/open-source), [#ai-agents](https://daily.dev/tags/ai-agents), [#mcp](https://daily.dev/tags/mcp)

[View this post on daily.dev](https://daily.dev/posts/the-opensource-managed-agents-hh62g6j2v)

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]}
{"@context":"https://schema.org","@type":"TechArticle","headline":"The OpenSource Managed Agents","url":"https://daily.dev/posts/the-opensource-managed-agents-hh62g6j2v","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/the-opensource-managed-agents-hh62g6j2v"},"datePublished":"2026-09-14T13:45:11.584Z","dateModified":"2026-10-15T02:22:13.681Z","description":"True Forge is an open-source, MIT-licensed, self-hosted agent execution harness from True Foundry that replicates the functionality of hosted 'managed agents'...","image":"https://i.ytimg.com/vi/3AdW345Jskk/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/3AdW345Jskk/sddefault.jpg","isAccessibleForFree":true,"articleSection":"Sam Witteveen","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Sam Witteveen","logo":"https://media.daily.dev/image/upload/s--gJm-KsgL--/f_auto/v1711727006/logos/samwitteveenai","url":"https://daily.dev/sources/samwitteveenai"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/the-opensource-managed-agents-hh62g6j2v","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"open-source,ai-agents,mcp","timeRequired":"PT20M","video":{"@type":"VideoObject","name":"The OpenSource Managed Agents","description":"True Forge is an open-source, MIT-licensed, self-hosted agent execution harness from True Foundry that replicates the functionality of hosted 'managed agents'...","thumbnailUrl":"https://i.ytimg.com/vi/3AdW345Jskk/sddefault.jpg","uploadDate":"2026-09-14T13:45:11.584Z","duration":"PT20M","url":"https://api.daily.dev/r/hh62g6J2V","embedUrl":"https://www.youtube.com/embed/3AdW345Jskk"}}
{"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Sam Witteveen","item":"https://daily.dev/sources/samwitteveenai"},{"@type":"ListItem","position":3,"name":"The OpenSource Managed Agents"}]}
{"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/the-opensource-managed-agents-hh62g6j2v#faq","mainEntity":[{"@type":"Question","name":"How much cheaper is True Forge compared to Claude managed agents for the same tasks?","acceptedAnswer":{"@type":"Answer","text":"True Forge cut costs by about 30% compared to Claude managed agents while maintaining the same accuracy, using the same Opus model. In a benchmark of 14 tasks from DevRev's enterprise bench, Claude managed agents used about 10 million tokens per run on average versus only 3.8 million tokens with True Forge. Switching to GLM 5.2 on True Forge brought total cost down 75% versus the original Claude managed agents setup. Developers weighing managed versus self-hosted agent harnesses can track cost benchmarks like this on daily.dev."}},{"@type":"Question","name":"What is True Forge and how does it differ from Claude or Gemini managed agents?","acceptedAnswer":{"@type":"Answer","text":"True Forge is an open-source, MIT-licensed agent execution harness from True Foundry that provides the same capabilities as hosted managed agent platforms (session handling, tool calls, sandboxes, approvals, streaming, a UI) but runs fully locally or in a private cloud with any model of choice, unlike Claude or Gemini managed agents which lock you into their own models and hosted infrastructure. Teams comparing self-hosted and vendor-managed agent runtimes can follow tools like this on daily.dev."}},{"@type":"Question","name":"How do I get started running True Forge locally?","acceptedAnswer":{"@type":"Answer","text":"True Forge is a TypeScript repo installable with a single npx command, which sets up a local SQLite database so all data stays on your machine, with hosted modes available using Postgres, Redis, and Docker. After installation, models, connectors (like Tavily or custom MCP servers), skills, and sandbox providers are configured through a settings UI, and agents can be tested via a built-in chat interface or called through an HTTP API and TypeScript SDK. Developers setting up self-hosted agent tooling can find practical walkthroughs like this on daily.dev."}}]}
```

