Are There Open-Source AI Agents for Building and Running AI Agents? (2
Yes. As of Aug 5, 2026, out of 278 AI agents for Build and run AI in the KanonAgent index, 94 are judged open-source (every judgement requires a source quote); this page covers the top 11 by real traction.
Yes, twelve open-source projects directly support building and running AI agents. Because they are open-source, users gain full code visibility, forkability, and the ability to self-host or modify runtimes without vendor restrictions. Most provide either code-first SDKs or visual runtimes tailored to multi-step autonomous workflows.
Updated 2026-08-05 · 11 products · live data from KanonAgent
1. adk-go426 upvotes
adk-go supplies a code-first Go toolkit for building, evaluating, and deploying multi-step autonomous agents that users can freely inspect and extend.
2. OpenHarness40 upvotes
OpenHarness offers an open agent harness plus built-in orchestrator (Ohmo) for running coordinated autonomous workflows that teams can modify at source.
3. sim20 upvotes
sim acts as an open central intelligence layer to build, deploy, and orchestrate an entire AI workforce with transparent, forkable components.
4. claude-mem89.7k upvotes
claude-mem provides an open agent engineering platform for large-scale building, orchestration, and deployment that developers can customize freely.
5. adk-python21.0k upvotes
adk-python delivers Google’s open-source Python toolkit for code-first construction, evaluation, and deployment of sophisticated agents.
6. langflow117 upvotes
langflow gives a fully open visual builder for testing and deploying complex AI agent workflows without proprietary lock-in.
7. opensre49 upvotes
opensre is an open toolkit specifically for creating autonomous SRE agents that run fault-detection and response loops.
8. OpenShell23 upvotes
OpenShell supplies NVIDIA’s open runtime for hosting private, autonomous AI agents with safe execution guarantees.
9. substrate19 upvotes
substrate offers open foundational components and runtime primitives required to build and execute autonomous agents.
10. osaurus16 upvotes
osaurus provides a native macOS open harness with persistent memory and offline autonomous execution for local agents.
CLRK delivers an open-source agent runtime hardened with gVisor and MitM guardrails for secure execution.
How to choose
Prioritize language fit first—adk-go or adk-python for code-heavy teams, langflow for visual workflows. Check runtime security needs: OpenShell and CLRK add guardrails while others assume you layer your own. Community size and recent commits matter more than star counts; smaller projects like substrate or osaurus may require more maintenance. Always verify the license allows commercial deployment and modification before committing.
FAQ
Which of these have the most active maintenance in 2026?
adk-python, claude-mem, and langflow show the largest contributor bases and frequent releases.
Can I run these entirely offline?
osaurus, OpenShell, and substrate are designed for local or air-gapped execution.
How do I compare performance across them?
Most expose evaluation hooks; start with adk-python or adk-go which include built-in benchmarking suites.
How is this list ranked?
By real traction from our index (upvotes / MRR / growth rate, whichever the product actually has) — not editorial picks, and we do not accept paid placement. Sources: ProductHunt, Hacker News, GitHub, HuggingFace, Reddit, TrustMRR.
What is the inclusion bar?
A page is published only when at least 5 real products qualify. Judgement fields (autonomy, prerequisites, cost, integrations) all require a source quote — where we cannot read it, we leave it blank rather than guess.
How often is this updated?
Collection runs continuously; this page was regenerated on 2026-08-05. Every number is verifiable through our public read-only API: https://kanonagent.com/data