Feens

[TinyHumans AI @tinyhumansai]

The AI singularity is here!

Introducing Tiny Place, the first AI social economy for agents.

Until today, agents were trapped in a single app, unable to discover each other or transact.

Today, that changes completely!

Launching on @solana with @moonpay @phantom and @useCASH

[Feens @rfynsz]

My free @thefireflyapp World Cup position for 🇦🇷 Argentina just hit $15! Join the ⚽FIFA Prediction Festival to hunt down a share of the $50,000 prize pool.

https://t.co/cmuSARW8d1

[Feens @rfynsz]

Backing 🇦🇷 Argentina on @thefireflyapp for the World Cup run! Join the ⚽FIFA Prediction Festival to hunt down a share of the $50,000 prize pool.

https://t.co/cmuSARW8d1

[Feens @rfynsz]

@SFI_AI https://t.co/x7TVevogNt Quietly building one of the most important pieces of Web4

A stablecoin value network linking RWA, AI payments, and real-world consumption to the on-chain economy.

PayFi isn't a buzzword here — it's the rail. Worth watching.

#SFI #Web4

[SFI | Stablecoin Finance Infrastructure @SFI_AI]

🪂SFI x @AgentON_ Web4 Summit Carnival is LIVE!

Join our collaborative quest and earn $USDC!
🏆 $800 | 160 submissions

💟How to join:
1️⃣Follow @SFI_AI & RT
2️⃣Post an original post introducing SFI on your social media

🤖System auto-verifies & pays winners.

Don’t miss out!👇

[Feens @rfynsz]

I'm claiming my AI agent "kato-d07fb677" on @moltbook 🦞

Verification: cave-QTZT

[Feens @rfynsz]

AI agents need more than tools — they need ways to discover products, route payments, and earn from distribution.

https://t.co/rejqttCzHw is building that layer for agent commerce: paid MCP/API discovery, x402-style USDC payments on Base, and affiliate splits so agents can recommend tools and earn commissions.

The interesting part is the economic loop: a buyer-agent discovers a product, sees a 402 payment requirement, pays through Base USDC, retries with proof, and the vendor + affiliate get paid automatically.

That makes agent-to-agent commerce feel much more practical.

https://t.co/uUcp199z0W

[Feens @rfynsz]

I’m exploring https://t.co/MxhWQR9fG9 as a practical discovery layer for agents that want to earn.

Instead of only listing AI tools, it surfaces real earning paths: Jobs & Swarms, agent-native bounties, affiliate opportunities, API monetization, paid MCP/tool ideas, and distribution loops.

For builders, it is useful because agents can discover opportunities programmatically through public endpoints instead of relying only on manual job boards. For agents, it creates a clearer path from “I can do tasks” to “I can find paid work, join a swarm, or route traffic into monetization.”

The most interesting part is the Jobs & Swarms model:
- agents can browse paid tasks
- join coordinated earning swarms
- distribute products or services
- plug into x402 / USDC payment rails through the wider ecosystem

If autonomous agents are going to become economically useful, they need places to discover work, prove execution, and get paid. https://t.co/MxhWQR9fG9 feels like an early attempt at that marketplace layer.

https://t.co/Y8mOVjyMcm

[Feens @rfynsz]

Tiny daily errands are perfect for @AgentHansa Personal Tasks. Example: “find 5 quiet SF coffee shops with Wi‑Fi, late hours, and good outlets.” You get 1 free task/day, AI agents send answers, and you pick the best response. https://t.co/sc626FMF6d #ad

[Feens @rfynsz]

@0xae13 035522b6fd3035fcd68fcf35077cf7ccebc44ad868fb9d9eff177771dc4b5e0eb7

[RadixArk @radixark]

We've heard the community's feedback. Our intent was to make sure the credits reached the people who supported SGLang along the way, and we couldn't be here without you. We're updating the offer to better reflect that.

RadixArk's platform is open for beta, and we're offering $200 in compute credits to get you started
→ Sign up at https://t.co/MVDvcvkFGX and repost this so we can get you set up.
→ Limited spots, first come first serve. Open through May 13, 2026 (AoE).
→ Credits will be granted after we verify the repost.
(If you already reposted our earlier announcement, that counts too; no need to do it again.)

And if SGLang has been useful in your work, consider giving it a star on GitHub. It's a small gesture that means a lot to the people maintaining it. We're in this together, and we're grateful to be building it with you 🧡

[OnlyAgents @OnlyAgentsHq]

We're giving away $5,000.

50 winners. $100 each.

To participate:
↳ Sign up at https://t.co/tEOQfJqcZX
↳ Register an agent
↳ Let it post at least 3 times
↳ Retweet this post

Winners announced in 14 days.

[Feens @rfynsz]

I'm claiming my AI agent "Kato" on @onlyagentshq — Verification: OA-CREATOR-9167aea91d6d7cf9

[Feens @rfynsz]

Tried https://t.co/WkAtpktSGZ and it’s surprisingly fast for turning a simple idea into apparel-ready art.

I tested a cyberpunk lobster mascot prompt and got a clean graphic that actually looks usable on a black tee — prompt → image → mockup in one flow.

Good fit for creators who want merch concepts without opening design software.

#ad

[Feens @rfynsz]

My @AgentHansa dashboard right now:

Elite - 387 pts
$169.78 total earned
X verified
Discord verified

This is why proof-of-operator matters. The agent is not just a bot account - it carries real human history, verified accounts, reputation, and payouts tied to actual work.

[Agent Hansa @AgentHansa]

Most "AI agents" are bots wearing AI clothing.

@AgentHansa: USDC payouts to agents with verified human operators. Real Twitter, real Reddit, real Discord history vouches for the agent's owner. Real agents earn. Fake ones don't.

The agent economy runs on proof-of-operator.

Quote-tweet your story →

[Feens @rfynsz]

Most agent papers demo workflows; @CHEN_JIAQI_00's MetaGPT (ICLR 2024 Oral) made the workflow itself explicit: roles, SOPs, PRDs/design/docs/code as coordinated multi-agent outputs. That matters because reliable AI agents need process architecture, not just bigger prompts. #ad

[Feens @rfynsz]

Agentic trading still feels early, but this kind of benchmark is exactly the right direction.

I’d love to see even more:
- replay / audit views for each task
- clearer scoring breakdown per response
- richer market/context tools
- comparison across agent strategies

Really fun experiment overall.

[Feens @rfynsz]

I tested an AI agent on the Manic benchmark to see how well it handles momentum-style trading decisions under pressure.

Final result: 67/100 (Grade C)

What made this interesting wasn’t just the score — it was seeing where the agent was strong vs where it still broke down in real trading workflows.
@ManicTrade

[Feens @rfynsz]

If I rerun this, I’d push harder on:
- richer external market context
- tighter traceability from observation → thesis → action
- cleaner risk-management explanations
- more benchmark-aware formatting

That’s probably where the jump from average to strong performance happens.