Pinned Tweet
kek will recap
all call
from start
$harambe @harambeonsolana
9k to 80m 8888x
$taoshi @taoshi_sol
when 6k too 20m
3333x
$bendog @ben_dog_
40k to 50m 1250x
$goo @Sillygooseonsol
6k too 20m
$orc @OrcSolana
14k too 15m 1079x
more good x will come
#solana #solanamecoin
xsusacc 🉐 retweeted
Qierion verifies one DAO decision executes consistently across connected chains $QIRO CA:
48bK3Ua5TacdfBJx3ABT7Z4QTcL3yfSJwWDbxNFKpump
xsusacc 🉐 retweeted
$NOCTA
CA: Eu9wsPurXKBcowasZwXUetkP7qMk8fe2B7svzCjpump
Nocta brings private infrastructure to AI agents on Solana.
Shielded USDC payments, private AI access, encrypted messaging and escrowed jobs. Powered by zero knowledge proofs and one time wallets.
usenocta.fun
$NOCTA
CA: Eu9wsPurXKBcowasZwXUetkP7qMk8fe2B7svzCjpump
Nocta brings private infrastructure to AI agents on Solana.
Shielded USDC payments, private AI access, encrypted messaging and escrowed jobs. Powered by zero knowledge proofs and one time wallets.
usenocta.fun
xsusacc 🉐 retweeted
SOLANA LAUNCH ROUTES
Pump.fun, Meteora DBC, or Raydium LaunchLab.
Pick a Solana launch path from Rivra and get creating.
xsusacc 🉐 retweeted
Start launching your own token, your own creativity with us.
Get the official $RIVRA token at @MeteoraAG
HdRrFTqEFf8HHLT8ejsjrYrsYEW323gm9x5JwbXrivra
xsusacc 🉐 retweeted
If you’re checking out Canny, start with the docs: what gets recorded, when “done” gets blocked, and how replay works.
The source is there too. Read it, question it, send me the awkward edge cases. That’s how this gets better.
canny-agents.cc/source
xsusacc 🉐 retweeted
GM. Experimenting with different agent structures today.
Starting with a single agent, then a planner delegating to specialists, then an executor paired with a separate reviewer.
Same task, same acceptance criteria. I want to see where the extra coordination helps and where it burns context, tokens, and time.
The hand-offs are what I’m most curious about. What does the next agent inherit? Can it see failed attempts? Who owns the current state when two agents disagree about what’s finished?
I also want to interrupt a few runs halfway through. Break a tool call, change a requirement, remove some context. See which setup can recover without needing me to reconstruct the whole situation.
Claude will be involved, obviously.
Hopefully by tonight I’ll have a better understanding of which tasks benefit from multiple agents. There’s also a chance I’ll have invented a very expensive group chat.
xsusacc 🉐 retweeted
Developer tokens have been burnt: solscan.io/tx/3AsGnYvnALJgjb…
new multi million runner today ?
github.com/qkal/Canny
$canny
6bDhqWwymacCxZKqRSvmCzCojN8eoLC4Ku5qjimbanny
#solana #meme #memecoin
After months of work, I’m proud to properly introduce Canny.
I’m also introducing the $CANNY token alongside the project.
CA: 6bDhqWwymacCxZKqRSvmCzCojN8eoLC4Ku5qjimbanny
This is my personal project and a brand I’m putting my name behind. A lot of time, thought and effort has gone into getting it here, and it means a lot to finally share it.
If you know me, you know I’m obsessed with Claude. That obsession also comes with wanting to understand exactly what happened between giving an agent a task and hearing “done.”
Canny is a local supervisor for Claude Code and Codex CLI. It records edits, commands and check results, and pushes back on completion claims when verification is missing after the latest code change.
The workspace brings that evidence together: inspect sessions, replay completion decisions, and test guardrails against commands and edits.
Being able to trace a result back to what actually happened is something I care about deeply. That’s the idea I’ve been building Canny around.
canny-agents.cc
I hope you like what I’ve been working on. Have a look, try it with your workflow, and tell me what feels useful or needs attention. I’m proud of this, and I want to keep making it better.
xsusacc 🉐 retweeted
After months of work, I’m proud to properly introduce Canny.
I’m also introducing the $CANNY token alongside the project.
CA: 6bDhqWwymacCxZKqRSvmCzCojN8eoLC4Ku5qjimbanny
This is my personal project and a brand I’m putting my name behind. A lot of time, thought and effort has gone into getting it here, and it means a lot to finally share it.
If you know me, you know I’m obsessed with Claude. That obsession also comes with wanting to understand exactly what happened between giving an agent a task and hearing “done.”
Canny is a local supervisor for Claude Code and Codex CLI. It records edits, commands and check results, and pushes back on completion claims when verification is missing after the latest code change.
The workspace brings that evidence together: inspect sessions, replay completion decisions, and test guardrails against commands and edits.
Being able to trace a result back to what actually happened is something I care about deeply. That’s the idea I’ve been building Canny around.
canny-agents.cc
I hope you like what I’ve been working on. Have a look, try it with your workflow, and tell me what feels useful or needs attention. I’m proud of this, and I want to keep making it better.
xsusacc 🉐 retweeted
Subscriptions are a state machine, not a payment. A technical tour of what recurring revenue in Web3 actually takes. 👇
xsusacc 🉐 retweeted
Introducing ArcenPay.com
The protocol for on-chain subscriptions!
Most Web3 products still sell access like it’s 2016.
A wallet in a @discord. A spreadsheet. A DM when something breaks. No invoice. No receipt. No audit trail. The product went on-chain. The billing around it never did.
AI agents made that gap worse. They can think, browse, and ship, but they can’t pay. No card. No bank. No standard way to prove a spend is authorized.
@arcenpayhq is the billing and entitlement layer that closes that gap.
We give teams one programmable stack to:
• sell subscriptions & usage in stablecoins
• gate features by real payment state
• settle across chains
• let humans and AI agents pay through the same rails
Not a checkout bolt-on. Infrastructure for selling access.
The problem we solve:
Running a paid product usually means wiring together: payment processor → subscription manager → entitlements → usage metering → invoices → customer ops → webhooks
Each piece drifts out of sync. Pricing changes become engineering projects. Access doesn’t follow payment. Agents get custom payment hacks that don’t scale.
ArcenPay connects those pieces into one layer:
Sell → Gate → Sync
1. Sell: embedded checkout, hosted payment links, agentic /spec, x402 micropayments, escrow
2. Gate: entitlements, feature flags, usage limits, credits
3. Sync: signed webhooks, invoices, lifecycle events
What can you build?
• Recurring subscription products
• Usage-based APIs
• Paid developer tools
• Digital memberships & token-gated access
• Credit packs & allowances
• Agent-paid services & machine-to-machine commerce
• Hosted payment links (human UI + machine-readable /spec)
• Hybrid wallet-native billing without card-first checkout
Built for SaaS, Web3 protocols, platforms, and AI/agent builders.
How it works (humans)?
1. Configure plans, add-ons, credits, and features in the provider dashboard
2. Embed billing with @arcenpay/react (ArcenPayProvider → identify() → ArcenEmbed)
3. Customer connects a wallet and activates a plan on-chain
4. Subscription state updates · entitlements unlock
5. Your app checks access with hooks or server-side APIs
6. Signed webhooks keep your backend in sync
Access follows entitlements - not the last checkout screen.
How it works (AI agents)?
Agents run on the same platform, without a human in the loop:
1. Fund a SessionVault with spend limits (maxAutoApprove, maxDailySpend)
2. Bound authority with ArcenAuth mandates (scope, chain, recipient, policy)
3. Discover pricing via payment-link /spec or hit a protected API
4. Receive HTTP 402 · sign a voucher · settle from session balance
5. Optional Verify-Then-Settle escrow - commit → verify deliverable → settle or auto-refund
x402 middleware in @arcenpay/node protects paid routes. @arcenpay/agent + MCP tools plug into @claudeai, @cursor_ai, and LLM stacks.
Same ledger. Same invoices. Same webhooks - whether a human or an agent paid.
Multi-chain settlement
Stablecoin settlement, one billing model:
@solana
@StellarOrg
@arc
@base
On-chain surface:
• SubscriptionRegistry - mint / renew / cancel subscription state
• ERC-7579 Autopay - recurring allowances without card rails
• PlanFactory - on-chain plan catalog
• FeeCollector - protocol fees
• SessionVault - agent balances, x402, escrow, revocation
On Solana, one Anchor program fills every role. Dashboard owns the catalog. Chain owns the subscription.
Non-custodial. Contracts settle. Platform manages state around it.
Entitlements & metering
Plans map to features, limits, and credits.
• React: useFlag, useEntitlement, useConsumeEntitlement
• Server: checkEntitlement, checkFlag, consumeEntitlement
• Feature flags with rules, plan auto-rules, and company overrides
UI can be optimistic. Anything that costs money should be enforced server-side.
Optional ZKVUB - zero-knowledge usage proofs so metered settlement doesn’t expose per-consumer volumes.
Developer surface
npm install @arcenpay/react # embed + gate in the browser
npm install @arcenpay/node # entitlements, webhooks, x402
npm install @arcenpay/agent # autonomous payments + MCP
Also:
• Stripe-style keys - sk_ · pk_ · rk_ · arc_tok_
• Hosted checkout - app.arcenpay.com/pay/…
• HMAC-signed webhooks (lifecycle events)
• Invoice builder + PDF delivery
• Environments (dev / prod) with separate catalogs
Docs: docs.arcenpay.com
API: api.arcenpay.com
Why teams choose ArcenPay
1. Wallet-native - billing designed for wallets, not card checkout adapted to crypto
2. Non-custodial - settlement stays on open contracts near your treasury design
3. Billing tied to access - plans and entitlements shape the product directly
4. Agent-ready - mandates, x402, escrow, MCP - not a human-only afterthought
Protocol economics are usage-based fees. Value from real settlement volume, not emissions.
Who it’s for
• SaaS & API companies shipping subscriptions / usage
• Web3 teams monetizing protocols, memberships, on-chain services
• AI builders who need agents that can discover, pay, and settle
• Platforms embedding billing & entitlements for their ecosystem
Status
Live settlement on production networks. Verified contracts. Dual-mode checkout (human + agent). Entitlement engine. Webhooks. SDKs. Litepaper with formal access & fee model.
We’re working with design partners who have real billing problems - not demos.
Start here
- Website: arcenpay.com
- Litepaper: arcenpay.com/litepaper
- Docs: docs.arcenpay.com
- GitHub: github.com/Arcenpay
- SDK: docs.arcenpay.com/sdk/react/…
- API: docs.arcenpay.com/api/overvi…
- Demo Video: youtube.com/watch?v=Y5o8oTCU…
- Telegram: t.me/arcenpay
- Article: xitter.cf/arcenpayhq/status/2084…
Every product sells access.
Every agent needs to pay for it.
ArcenPay connects the two.
Book a demo from the site
$ARCENPAY is now live on solana:So11111111111111111111111111111111111111112 via @Pumpfun
CA: AGxb3YSQAQbd8iPx9SFucG6vWx2NYQ5Qfht493CVpump
🤖 Made with AI
Biggest Update: Product Video is Live!
Wallet-native subscriptions. Entitlement gating. Agent payments. Settled on-chain.
Watch how ArcenPay actually works ↓
youtube.com/watch?v=Y5o8oTCU…
The protocol for on-chain subscriptions.























