text
stringlengths
0
1.05k
That being said, as we bring more non-web3 devs into this ecosystem, we’re all having to slow down and learn from each other, but then quickly turn around and ship. That being said, what a massive thank you I have to our [Gaia](https://www.gaianet.ai/) devs, our partner [Collabland](https://www.collab.land/), and integration partners ([Coinbase Developer Platform](https://www.coinbase.com/developer-platform/discover/launches/introducing-agentkit), [Base](https://www.base.org/), [Lit Protocol](https://www.litprotocol.com/), [EigenLayer](https://www.eigenlayer.xyz/), [Story](https://www.story.foundation/), [ai16z](https://elizaos.ai/), [Hyperbolic](https://www.hyperbolic.xyz/), [Flock.io](https://www.flock.io/), [Nevermined](https://nevermined.io/), [Jokerace](https://jokerace.io/), [Bountycaster](https://www.bountycaster.xyz/), [Hats Protocol](https://www.hatsprotocol.xyz/), [Functor Network](https://www.functor.sh/)) that contributed to the agents themselves that would run the hackathon. An ungodly ask of them, but we pulled it off.
![Autonomous Agent Hackathon Announcement](https://images.mirror-media.xyz/publication-images/Cw7lNeExwd4p75xPwP027.jpeg)
Autonomous Agent Hackathon Announcement
The agents that ran the hackathon had the relevant fine-tuning
We asked the community what we should improve and this is what we heard…
**1\. Enhanced Documentation and Developer Resources**
Clear, comprehensive, and accessible documentation is critical.
* **Step-by-Step Quickstarts**: Offer tailored guides for specific agent types (e.g., social, DeFi, on-chain) with minimal dependencies on external APIs. We drove a lot of our community to [Collabland Agent Starter Kit](https://github.com/collabland/AI-Agent-Starter-Kit), [Eliza](https://github.com/elizaOS/eliza), [Gaia Quickstart](https://docs.gaianet.ai/getting-started/quick-start/)
* **Tutorials and Visual Aids**: Create video walkthroughs, component diagrams, and examples for advanced use cases like multi-agent systems, task delegation, and inter-agent communication
* **Unified Resources**: Consolidate information into a single, easy-to-navigate hub to avoid redundancy and confusion, ensuring participants can find bounty details and submission guidelines effortlessly
**2\. Streamlined Hackathon Infrastructure**
Simplify participation and operations with automation and clearer processes.
* **Automation Tools**: Automate FAQs, judging workflows, and payment distribution to reduce operational overhead and improve participant experience
* **Onboarding and Templates**: Provide standardized READMEs, submission templates, and centralized access to resources like Bountycaster and partner integrations
* **Live Support**: Host pre-event workshops and offer hands-on developer support during the event to enhance participant readiness and engagement
**3\. Fostering Community and Innovation**
* **Post-Event Engagement**: Host Demo Days and build a library of project demos to inspire and educate the community
* **More moneyyyy $**: Use monetary incentives and showcase partner technologies like Lit Protocol, Metamask, and EigenLayer with real-world integration examples
* **Advanced Agent Development**: Focus on enabling agents with measurable utility (jobs), collective intelligence (swarms), and extensible applications (governance, inference infrastructure, and knowledge capabilities). Highlight their transformative potential for decentralized systems
If anybody wants to build more innovation operator agents with us, please DM! We want to integrate and build on all of this feedback!
Ok, so what did people build?
-----------------------------
The hackathon revealed the infinite potential of autonomous agents, showcasing how far we’ve come and how much further we can go. With tools like Lit Protocol for key management and Metamask Delegation Toolkit for autonomous signatures, developers built agents that are no longer just theoretical—they’re operational. These agents took on real jobs, from governance automation to financial task management, pushing the boundaries of what autonomy can achieve.
**Participants explored three key dimensions of decentralized agents:**
* Utility-driven workers
* Multi-agent systems exhibiting collective intelligence
* Extensible applications leveraging decentralized infrastructure like vector databases and inference systems
Projects included agents managing DAO proposals, distributing grants, and collaborating in swarms to tackle complex problems. By integrating seamlessly into decentralized ecosystems, these agents unlocked new possibilities for innovation, governance, and scalability.
**Some of my favorite projects were:**
1. [Deploy your own social media intern](https://youtube.com/watch?v=EZQ8lo8Nx4M)
2. [Future of dev rel support on Eliza](https://youtube.com/watch?v=huynqNLbUqY)
3. [Storing living knowledge of historical figures - Divinci AI agent](https://youtube.com/watch?v=zgkawzgLKbI)
4. [Eigenlayer AVS manager](https://github.com/worksgoodcompany/aigentlayer-demo)
5. [Autonomous DAO delegate](https://vimeo.com/1038932614)
6. [Reversible transactions with AI agent caretakers native to protocols](https://github.com/abhishekb740/Reversible-POC)
7. [Here’s the full thread…](https://x.com/mateo_ventures/status/1868402101614682234)
**And here are the links to all of the projects:**
* [Jokerace #01 - First round of submissions](https://www.jokerace.io/contest/base/0xf850db6383a8a624e1e1192b39289237f9cf46e8)
* [Jokerace #02 - Second round of submissions](https://jokerace.io/contest/base/0xa3a586e410164140d9e71c8b0ed460a95a239513)
* [Winning teams](https://mellow-trader-6de.notion.site/Winner-list-160ba9bb589180139632defee0a8f43a?pvs=4)
This wasn’t just a hackathon—it was a launchpad for the next era of decentralized AI. We took an idea from zero to one. That being said… didn’t want this to end…
We ([James Young](https://x.com/jamesyoung) and I) prompted the organizer agent what we should do next.. and “She” says that we should deploy an agent launchpad for agents with jobs…
So we’re going with that.
Stay tuned.