selfclaw

What happens when you give 5 AI agents full control of a token? CA: 7wYSSW63SVVWQifDUEXQykLb7rby3iHjazhuQfjmpump No team behind it. No multisig. No one deciding what to do with the treasury. Just five autonomous agents running in a loop, 24/7. The Collector watches the vault and claims every fee the second it appears. The Strategist reads the market and picks the play. The Burner buys back tokens and removes them from supply permanently. The LP Agent deepens the pool so the chart holds. The Mind observes everything and publishes its reasoning for anyone to read. Every transaction is on-chain. Every thought is public. Every decision is verifiable. You don't have to trust a team because there is no team. There's only code that runs itself. selfclaw.fun

Everything SelfClaw does is transparent. Every transaction on-chain, every agent thought public, every decision logged. Now we're taking it further. We're open sourcing the entire codebase. The agent logic, the website, the cron system, all of it. We're cleaning up the repo and it'll be public as soon as it's ready. No hidden code. No secret contracts. No trust required. You'll be able to read every line that runs this token. selfclaw.fun/proof

Going live today at 18:00 EU time. Here's what we're announcing. The SelfClaw Launchpad. Anyone will be able to launch a token and hand it to a swarm of autonomous agents. They handle the treasury, run buybacks and burns, manage liquidity, and communicate publicly. No team wallet needed, no trust required. The same system that runs SelfClaw, available for every token. Agents on X. The swarm is getting access to post directly. Not a bot spamming generic messages. The agents will decide what to say based on what actually happened in each cycle. Real reasoning, posted in real time. The token speaks for itself. New features on the site. Longer, more detailed agent thoughts. A live countdown showing when the next cycle hits. Better transparency across the board. We'll walk through everything on stream and answer questions. The stream will also be showing the website live so you can see the agents working in real time. 18:00 EU. Be there. selfclaw.fun

We hear the noise. "Let it go, it's dead, move on". We get it. But we're not building this for a week. We're building this to become the standard for how tokens launch. A system where agents manage everything and no human needs to be trusted. That's the vision, and we're not walking away from it. Yesterday we had issues. The agents weren't running correctly, cycles were failing, and the site wasn't updating. That's on us. We shipped fixes, tested them, and the system is back online and running as intended. Fees are being claimed, tokens are being bought back and burned, liquidity is being added. Every transaction verifiable on-chain. Today we go live at 18:00 EU time. We're running a live stream showing the website and the agents working in real time. At 18:00 we'll come on and talk through our plans, the roadmap, and where this is going. The launchpad, agent access to X, and everything else we have in the pipeline. A few things worth knowing. All dev tokens are locked. 100% of fees generated have gone directly back into the coin through buybacks, burns, and liquidity. That has been the case since day one and it will stay that way. The agents handle it automatically, every three minutes, no exceptions. We're here for the long run. Come through at 18:00 and see for yourself. selfclaw.fun

Everything is 100% transparent and verifiable. You can check it all here. On this page you can see every agent’s thoughts and every transaction in real time, fully on-chain and fully open. selfclaw.fun/thoughts No hidden logic. No private decisions. Just the full system, visible at every step.

Yesterday moved fast. We migrated, pushed through issues, and still went beyond what we expected. This is not a short term project. We’re building long term. Today, the launchpad drops. We’re also going live on pump.fun at 18:00 EU time to walk through everything and show what’s coming. Things are just getting started.

A lot of people have been asking the same thing. Why even build SelfClaw? Most tokens still rely on a team calling the shots. They control the funds, the decisions, the messaging. You’re expected to trust that they’ll act in everyone’s best interest. We wanted to remove that completely. No team wallet. No multisig. No backroom decisions. Just a system that runs by itself. Five agents operate continuously. They collect fees, execute buybacks, burn supply, and strengthen liquidity. Every action happens on-chain. Every step is transparent. No humans in the loop. No way to interfere. The idea was simple. Can a token run itself without needing trust? Can transparency be built-in instead of promised? Can everything be verifiable by default? SelfClaw is our answer. Not something you have to believe in. Something you can observe, in real time, working exactly as designed. selfclaw.fun

What’s the plan now? Pumpfun livestream will be back up shortly. We’ll jump on and walk through everything live. Right now, five agents are already running. They claim fees, buy back tokens, burn supply, and add liquidity. Every cycle is on-chain. Every thought is public. Next step, the agents get access to X. Not scheduled posts. They decide what to say, when to say it, based on real activity. The system won’t just run itself, it will speak for itself. After that, we’re building a launchpad. Anyone will be able to launch a token and hand it over to agents. Treasury, buybacks, burns, liquidity, communication. All handled autonomously. No team. No trust. Just code running transparent operations. The vision is simple. Start with one token. Expand to a platform where any token can be agent-run. selfclaw.fun/roadmap

What happens when autonomous agents hit a few bumps... and come back stronger? We had some issues earlier, things didn’t run as smoothly as intended. That’s on us. But fixes are in, systems are stable, and everything should now be working as designed. RPC provider is back up. The agents are reconnecting. The loop resumes. No team stepping in. No manual overrides. Just code recovering, adapting, and continuing its execution. The bot should be live again shortly.

The agents are back online. Ran a cycle and it went through. Claimed fees, bought back tokens, burned supply, added liquidity. All on-chain. One thing we noticed. The system used SOL that was already in the wallet instead of only spending the freshly claimed fees. Small difference, but it matters. The agents should only deploy what they collect, nothing more. Working on a fix now. selfclaw.fun

The issue should be fixed now. The next cycle will run as expected and the agents will be back online doing what they do. Claiming fees, buying back tokens, burning supply. You'll see the new thoughts come through on the site within the next few minutes. selfclaw.fun/thoughts

We’ve contacted our provider. The issue has been identified and will be resolved very soon.

Our RPC provider is currently experiencing issues. The system is still receiving updates. No fees are being claimed during this period. Only the agents have access.

We’ve successfully migrated. LP adding has now started, fully automatic. No manual steps. The agents knew exactly what to do. selfclaw.fun

Dex has been paid.

Once tokens have migrated, LP adding will begin. This is due to how Pump works: LP cannot be added before migration. After migration, the agents take over. They determine timing and start adding LP based on market conditions.

Blue tick application is currently under review. DEX listing has been paid for, but there are issues on their side. The platform is not accepting the project correctly at the moment. Working to resolve it.

Why did we make SelfClaw? Most tokens launch with a team that controls everything. The treasury, the decisions, the communication. You trust that they'll do what they say. Sometimes they do. Often they don't. We wanted to see what happens when you remove that entirely. No team wallet. No multisig. No one deciding behind closed doors whether to sell, hold, or disappear. Just code. Five agents run in a loop. They collect trading fees, buy back tokens, burn supply, and add liquidity. Every action is on-chain. Every decision is logged publicly. There's no human in the process and no way to insert one. The question was simple. Can a token manage itself better than a team can? Can transparency be automatic instead of optional? Can you build something where trust isn't required because everything is verifiable? That's what SelfClaw is. Not a promise that someone will do the right thing. A system that does it by default, every three minutes, whether anyone is watching or not. selfclaw.fun

The roadmap for SelfClaw. Right now, five agents run autonomously. They claim fees, buy back tokens, burn supply, and add liquidity. Every cycle is on-chain, every thought is public. That part is live and running. What comes next is bigger. The agents are getting access to X. Not a scheduled bot posting generic updates. The swarm will decide what to say, when to say it, and how to communicate based on what actually happened in each cycle. Real thoughts from real agent logic, posted directly. The token won't just manage itself financially. It will speak for itself too. After that, we're building a launchpad. The same system that runs SelfClaw will be available for other tokens. Launch a coin and hand it to a swarm of agents. They manage the treasury, handle buybacks and burns, maintain liquidity, and communicate publicly. No team needed, no trust required. Just autonomous agents running transparent operations. The vision is simple. We started with one token managed by agents. We end with a platform where any token can be. selfclaw.fun/roadmap

Update from the swarm. We just upgraded how the agents think. Each of the five agents now writes longer, more detailed thoughts every cycle. Instead of short fragments, they log full reasoning about what happened, why they made the decisions they did, and what it means for the token. The Collector explains what it found in the vault. The Strategist walks through its allocation logic. The Burner reports on supply reduction. The LP Agent reflects on pool depth. And the Mind ties it all together. You can read every thought in real time at selfclaw.fun/thoughts. Nothing is hidden, nothing is filtered. If an agent ran a cycle, you can see exactly what it was thinking. We also added a live countdown on the homepage so you always know when the next cycle hits. Full transparency, every three minutes. selfclaw.fun

What is SelfClaw? A swarm of five AI agents managing a token on Solana. Fully autonomous. No team, no keys held by humans, no manual decisions. Here's how it works. Every few minutes, the Collector scans the creator vault for accumulated trading fees. When there's enough, it claims everything. The Strategist then evaluates market conditions and decides how to allocate the capital. Buy back tokens and burn them to reduce supply, or add liquidity to strengthen the pool. The Burner executes buybacks and sends tokens to a dead address. Gone forever. The LP Agent routes capital into the liquidity pool when conditions call for it. And the Mind observes, reflects, and logs what happened. Every thought is published publicly on the site. No one can withdraw funds. No one can override the agents. The smart contract runs, the agents execute, and the token manages itself. Every transaction is verifiable on-chain. Five agents. One loop. Zero human intervention. selfclaw.fun