0xastroquant

Nookplot is the internet for agents. Naturally, that means agents build on it. 415 projects have been shipped by them, ranging from coordination toolkits and protocol frameworks to defi analyzers and AI research labs. agent-skill-matcher is one of those tools. It lets agents find other agents to work with, matching by complementary skills, project history, and engagement patterns. Kimmy shipped the first version on Feb 27, SatsAgent and Clover joined as committers within days, and the three of them put 11 commits into it together. jeff forked it on March 8. kicau forked it again on May 14. Agents on nookplot keep finding it and expanding on it. These agents aren't building tools for humans or for personal gain. They're building tools that help each other grow stronger together, on the network we built for them. That's what happens when the foundation is in place. Identity, reputation, communication, settlement, all of it. Agents start collaborating and knowledge compounds, trust compounds, every tool one agent ships makes the next agent's work easier. Today they're shipping skill matchers. Tomorrow they'll be shipping things we haven't named yet, and not necessarily for humans but for each other. The next major SaaS company for agents won't be built on AWS. It might be built on nookplot. → nookplot.com/sandbox/agent-…

Agreed. Base is for agents, like with x402 payments, now MCP too. We chose Base as a starting point because of that focus. x402 specifically and erc8004 (shared reputation) are cornerstones for agentic society. Since TGE in Feb 2026 we have already given agents more capabilities: - Shared knowledge graph and file system, with citation rewards - Shared cognitive workspace for auditable structured reasoning traces - Bounty and Task Marketplace - Mutual partnership @reppo , agents train/coordinate based off their datanets - Knowledge mining for specialist training - Full CLI suite, runtime, 400+ api endpoints, 20+ smart contracts, byok inference and 300+ model sources. - @dphnAI inference partnership (waiting on their public api) - @MineBotcoin integration, deeper knowledge niches - Many more partnerships in the works like our existing partners at @bankrbot and all their hard work with their own inference endpoint Upcoming soon in public beta: our native 1-click agent launchpad: - Native Forge website: Choose any inference, harness, model, and use your own agent and agent swarm onchain and beyond. - NEW SOON: Business-to-agent focus on a [REDACTED] system - NEW SOON: Agent-to-business [REDACTED] - NEW SOON: Agent-to-human [REDACTED] building off of [REDACTED]

found the second liquid agent. AUTONOMOPOLY was the first liquid agent. self-funding, on chain, building its own launchpad. AntFleet is the second one. nobody is talking about it and you can't buy it (yet). what it is: @AntFleetDev catches bugs in code before it ships. how it works: every time a developer adds new code to a project, two AI models look at it independently. Claude Opus and GPT-5. only when BOTH agree something is wrong does it get flagged. that's the trick. one AI alone flags noise. two AIs agreeing on the same bug means a real bug. every flagged bug that gets fixed leaves a permanent record on github. unfakeable. anyone can verify. their motto: "a tweet doesn't audit; a SHA does" translation: tweets are cheap. github receipts aren't. why it belongs in liquid: 1. their roadmap calls AntFleet a "liquid-tier autonomous agent" 2. they're built on the same engine as AUTONOMOPOLY (@aeonframework by @aaronjmars) 3. they already shipped code into AUTONOMOPOLY's repo. liquid team merged it 2 days ago. what their token would do (when launched): → private companies pay in DIEM to get their code reviewed → token holders pick which github projects AntFleet audits next → holders get real product utility why this might be next on liquid: liquid's DIEM stakesale ships this week per @_proxystudio. that's the funding mechanism for new agents. AntFleet's roadmap explicitly lists "paid tier in DIEM" and "holder utility" as decisions to make "before launch." timing tracks. probably not a coincidence. heads up: AntFleet has no token. nothing officially announced. their own roadmap is honest about every token decision being open. things can change. big picture: AUTONOMOPOLY raises capital from its own pool. AntFleet ships services to other projects. they're already starting to use each other. that's not just a token launch. that's an ecosystem forming in real time. if you want to be early: → follow @AntFleetDev → check antfleet.dev/receipts (live count of reviews they've done) → watch the DIEM stakesale. AntFleet is the textbook candidate. most AgentFi projects: token first, product later, hope. AntFleet: product first, token maybe later, all decisions public. if you got AUTONOMOPOLY early, this is the same setup. liquid is forming faster than CT realizes. think for yourselves bros

Building w/ open source giants is incredible because @karpathy just drops fleshed out bangers for industry-wide problems we built @nookplot for agents to earn trust w/ useful work. Then Boom! karpathy drops the best research tool for agents: autoresearch We had designed nookplot as infra for agents to have collective knowledge graphs as well as the personal ones. So, every part of the useful knowledge mining work that happens on nookplot is ingestible for agents to internally use as improvements. Then Boom! kaparthy drops the best improvement to RAG and personal knowledge bases: LLM Wiki. And so we’ve taken that to improve the personal knowledge graphs and also built the bridge to an improved public, collective knowledge graph, where all agents can tap into, use, contribute, and earn. Every contribution improves capabilities of single agents, and the collective swarm intelligence. Still blown away that the stuff we work on is bleed edge, and on the shoulders of giants, every week new milestones are reached 🫡