Hey Product Hunt!
We ve been running agent workflows for years, long enough to know exactly how they break. One agent makes a wrong assumption, another takes it as fact, the notes absorb it, and the error compounds. More instructions don t fix that. Seeing what actually happened does. So we built guilds.run: agent teams you can understand, debug and control.
Get started in a few minutes
Run it on your machine: git clone <https://github.com/CigaleAI/guil...> && cd guilds && ./setup.sh && docker compose up --detach --wait
Open http://127.0.0.1:8080, connect your LLM provider, and set up a guild of agents, each with its own task and model.
Give them a real task and inspect what happens: the exact prompt, every tool call and result, every memory change, and the group chat between agents.
When something goes wrong, ask Studio why. It reads the runs, finds the cause and drafts a fix. Nothing changes until you publish.
Hey Product Hunt! 👋
Getting AI to handle long-running tasks reliably takes more than good prompts. There’s a lot of system design involved: figuring out how agents should work together, anticipating edge cases, and being able to debug and tune the system when things go wrong.
That’s why we built guilds.run.
The goal is to make the whole system understandable, so you can trace how mistakes propagate between agents, fix the underlying causes, and see if your changes work.
We’re open-sourcing guilds.run because we’d like to see what other people built and share working team the solve real world use cases for long running agent collaboration
Coming next:
🧩 Recipe sharing: discover, reuse and share whole agent teams built by the community.
☁️ Hosted mode: the same product without managing your own infrastructure. Leave your email on guilds.run to get a seat when it opens.
If you like what you see, a ⭐ on GitHub helps a lot. We’re here all day, so tell us what breaks and what you’d build.
@m4nuc startups building AI powered products, reliability becomes increasingly critical as workflows grow more complex. Utilizing tools that simplify the process of monitoring and improving agent behavior can save teams significant time otherwise spent on debugging. Congratulations on the launch of the new product!
@angelaaa Thank Angela.
Congrats on the hunt @kate_ramakaieva
I've been using Grok Bot to monitor X for trends, viral topics, product mentions, and relevant conversations. Could I use guilds run to coordinate multiple agents across X, Reddit, LinkedIn, and Instagram? I want to have a spreedsheet with mentions of the topics/mentions etc.
@kate_ramakaieva @byalexai Thanks Alek,
Absolutely, this sounds like a complex use case and guild is the right tool for that. We are actually planing on publishing "templates" very soon, with usecases similar to what you describe.
Also I would love to hear about what you wish Grok Bot did better ? Are you happy with your trend monitoring ? How long have you been using it ?
@kate_ramakaieva @m4nuc That's exactly what I'd love Grok Bot to do!
Find relevant mentions across Reddit and Instagram, organize them in a Google Sheet/Notion, and send me reports in Telegram.
Grok Bot already does it in X, but Instagram is not possible. I'd love to have the same capabilities across Reddit and Instagram.
But Instagram's API doesn't support broad keyword searches and Reddit has tightened its API access. This is why I am looking for solutions.
@byalexai @vincent_cigale @m4nuc @kate_ramakaieva For sure! We'll post updates as templates evolve. If you have specific use cases in mind, we'd love to hear them 🙌
Hey PH! 👋
Super happy to be hunting guilds.run today.
Anyone who has run multi-agent setups knows the pain. One agent makes a wrong assumption, the next one treats it as fact, and 3 steps later you’re staring at a broken output with no idea where it went sideways. Most tools just hand you better prompts. guilds.run actually lets you see what happened.
What caught my attention:
🔍 Full visibility into every prompt, tool call, memory change and the group chat between agents
🛠 Studio reads your runs, finds the root cause and drafts a fix, and nothing goes live until you publish it
⚙️ Per-agent control over permissions, context and model
🆓 Open source and free to self-host, up and running with one docker command
Love that the team is building in the open.
Congrats @m4nuc and @vincent_cigale on the launch!
@vincent_cigale @kate_ramakaieva Thanks Kate. Great summary. We might keep it
self hostable + you can see every tool call, thats what makes agents trustable imo 🙌 any plans for a hosted version?
Hey Petr @petrkovacik, glad you feel so, I personally need to see everything that's going on before trusting them with meaningful tasks.
Hosted: Yes, it's coming soon too.
@petrkovacik +1! And if you don't want to miss the hosted launch, follow us on X (@guildsrun). updates dropping there soon 🚀
As more non-developers start building with AI, we’ll need ways for them to understand and debug the systems they create. Interested to see where this goes.
@dipanshu_kushwaha5 thanks for your comment
You've got it , that's exactly why we built guilds.run :)
Can I test a workflow safely before letting the agents run it for real? good luck!
@rania_rimali Absolutely, my cofounder does that all the time when creating workflows.
Thanks Rania
interesting product! how much does the hosted version costs?