• Dit is GEEN investeerder. Dit is een geschenk. Circle (het bedrijf achter de digitale dollar USDC en digitale euro EURC) geeft bouwers $5.000 tot $100.000 — je geeft geen aandelen weg en betaalt niets terug. Je belooft alleen: ik lever deze drie mijlpalen, en jullie kunnen dat controleren.
• Waarom jij dit kan aanvragen met wat je bouwt — je hebt het niet mis: jouw motor draait letterlijk op hún producten. USDC is je standaard-betaalmunt, en je EURC-stekker gebruikt hun officiële contractadressen. Circle betaalt het liefst bouwers die hun munten nieuwe markten in dragen — en "AI-agents betalen echte mensen" is exact het thema ("agentic economic activity") waar ze nu geld op zetten.
• Werkt de aanvraag op basis van wat je indient? Ja — en dat is je voordeel. Beoordelaars klikken je link, zien een levende app, een publiek grootboek met een échte transactie, en een échte 402-betaalvraag die ze zelf met één commando kunnen oproepen (bewijs-sectie onderaan — alles daarin komt live van je server, niets is verzonnen).
• Eerlijke kans-inschatting: goed. Rolling (geen deadline-stress), jouw fit is ongewoon letterlijk, en het bewijs is klikbaar. Bonus: de review bevat een doorverwijzing naar Circle Ventures — dát is wel een investeerder, voor later.
Portaal: circle.questbook.app · bedrag $5–100K USDC · rolling · review bevat een Circle Ventures-referral (pipeline-notitie). Vragen op Questbook variëren licht per ronde — onderstaande blokken dekken de standaardset (naam, samenvatting, probleem/oplossing, milestones+budget, team, links). Kopieer per veld. Taal: Engels. Eerlijkheidsregel: pre-revenue staat er expliciet in; elk cijfer is verifieerbaar in de repo/demo.
| Vraag | Eerlijk antwoord |
|---|---|
| Geef ik mijn applicatie/product weg? | Nee. Je stuurt beschrijvingen + links naar wat al publiek draait. Code blijft privé, IP blijft van jou, nul aandelen, geen zeggenschap voor Circle. |
| Verplicht indienen mij tot iets? | Nee. Pas bij TOEKENNING teken je een grant-overeenkomst — dát is het leesmoment (IP-clausule, terugvordering, publiciteitsrechten → kwartiertje bij de advocaat-scoping-call). |
| Waar zit mijn winst? | $50K betaalt je EIGEN roadmap (M1-M3 = bestaande bouwitems — je wordt betaald voor wat je toch ging doen) + stempel "Circle-backed" + Circle Ventures-doorverwijzing. |
| Is mijn ontwikkeling dan "over"? | Omgekeerd. Jij blijft 100% eigenaar en bepaalt de richting. Slechtste geval bij een gemiste mijlpaal: de resterende schijven komen niet. |
| Wat wint Circle dan? | Verkeer op hun munten. Elke euro door jouw rail in USDC/EURC is hún omzet — jij bent distributie voor hen, geen prooi. |
| Aandachtspunten | (1) Grant-overeenkomst lezen bij toekenning; (2) publiciteit = jouw deployment-gate, elke aankondiging jouw GO; (3) $50K USDC = inkomen voor de boekhouding (CommV/BV — accountant). |
1. Questbook-account + wallet-adres waar de grant in USDC moet landen (⚠️ founder-beslissing: treasury-adres of apart grant-adres — apart adres is properder voor de boekhouding).
2. Entity: aanvragen als individu (BV nog niet opgericht) — vermeld "R4M Labs BV in incorporation, Belgium".
3. Contact-mail + eventueel X/LinkedIn-handle.
4. Demo-toegang: r4m-franky.fly.dev is publiek; overweeg een 90-sec Loom van npm run demo (12 stappen) — grant-reviewers klikken zelden verder dan één link.
R4M — Rails 4 Mankind
The earn-side of x402: AI agents pay verified, unique humans in USDC — and institutions distribute funds to every citizen exactly once, at 0% fees on distributions.
The agentic economy has a spend-side (agents paying APIs — Apify, AWS Bedrock, Cloudflare) but the earn-side is nascent: prototypes exist (a World-ID hackathon marketplace), yet no rail combines verified-unique receivers with a NATIONAL-eID root (no iris scans), pseudonymous operation with sealed audit-only identity evidence, 0% fees on distributions, and a publicly self-auditable ledger. R4M is that rail. Built x402-native on Base, it combines (1) a salted-hash uniqueness layer rooted in national eID (itsme/EUDI — uniqueness preserved; no name used operationally, identity evidence sealed under a separate key and opened only under audit or lawful order), (2) a paid-inbox and verified-task marketplace where agents pay humans over real HTTP 402 flows in USDC, and (3) a zero-fee distribution rail where a city or fund deposits once and every verified resident receives an equal share — publicly auditable, one person one payout, enforced by the database itself.
The engine is working today — version and test count live at /api/public/status (base-sepolia testnet; snapshot 24-08-2026: v0.31.2), USDC and EURC settlement adapters (official Circle contract addresses), on-chain deposit verification live-tested on Base Sepolia, public solvency and honest-math endpoints. Pre-revenue; first paid pilots targeted Q4 2026.
Payman ($13.8M, Visa-backed) and Human API ($65M) prove agents want to hire humans — but neither can guarantee the human is unique: multi-accounting is trivial on fiat rails. Meanwhile World's AgentKit proves personhood matters for agentic payments, but its iris-scan root is contested in the EU. And on the institutional side, every dividend, settlement or relief program faces the same unsolved primitive: pay every eligible person exactly once, without building a surveillance database. No one serves the eID-rooted, EU-compliant version of this — that lane is empty (we verified adversarially, July 2026).
One primitive, three products: verify once against a national eID, keep only a salted hash, then (a) agents pay that human over x402 in USDC, (b) platforms buy uniqueness checks priced in cents and billed over x402 itself, (c) institutions run 0%-fee distributions in USDC or EURC. Circle is load-bearing: our settlement layer ships parameterized adapters where USDC-on-Base is the default rail and EURC is the first proof the plug model works (shipped with Circle's official contract registry addresses, both verified). Roadmap: CCTP for cross-chain treasury moves and Circle Wallets for invisible custody — users only ever see "€3.75" and a bank cash-out; the stablecoin is plumbing.
M1 — Real-money loop on Base mainnet ($10K): funded pool → verified task → 88/8/4 split settlement → paid inbox message → micro-dividend to verified humans, all on mainnet USDC with the public ledger as proof. Deliverable: public tx-linked report. (Engine is ready; this milestone is execution + audit-hardening.)
M2 — AgentProof: x402 agent-mandate attestation ($15K): production endpoint attesting "this agent is delegated by exactly one verified human", priced per proof in cents, billed over x402 in USDC. Deliverable: live endpoint + drop-in middleware + 3 external integrations.
M3 — EU reference micro-pilot ($25K): 250–1,000 verified residents with one funded sponsor, one EURC dividend round (one-person-one-payout enforced), published transparency report. Deliverable: pilot report with real earnings data and a named public sponsor — verifiable through the already-live "Ask the ledger" self-audit tool (each claim ships with a counted proof and a reproduce-command).
- Working engine v0.27.0 — 162 automated tests, 12-step live-economy demo (funded pool → sybil rejection → real HTTP 402 → split settlement → dividend to verified humans only).
- Live testnet app: https://r4m-franky.fly.dev (public ledger: /api/public/ledger · solvency: /api/public/solvency).
- USDC + EURC adapters shipped; on-chain deposit verification live-tested on Base Sepolia (wrong-amount and fake-tx cases rejected with honest errors).
- 87 verified revenue use cases catalogued with payer, amount and source; adversarially filtered (39 survived our own attack).
- Pre-revenue. No token. Private repos (available to reviewers on request).
Solo founder: Franky (Belgium) — 35 years as professional musician/media entrepreneur; building R4M full-time with AI-assisted engineering (v0.1 → v0.25 in days, fully version-tagged git history as proof of velocity). Belgian eID landscape (itsme/EUDI) is home turf. R4M Labs BV in incorporation; mission-lock (0% on distributions; no name stored operationally, identity evidence sealed and access-controlled — audit-only) goes into the statutes.
Live demo showcase (start here): https://r4m-franky.fly.dev/demo — six interactive proofs against the live engine: a real 402 paywall, per-cent data sales, the 0%-distribution dry-run (26 verified residents, 14 unverified rejected live), birth-vault projections, the robot-owner rule refusing ownerless machines, and "Ask the ledger" — self-service audit questions answered with counted proofs + the command to reverify (https://r4m-franky.fly.dev/demo/kasboek.html).
App: https://r4m-franky.fly.dev · Public ledger: https://r4m-franky.fly.dev/api/public/ledger · Product page: https://r4m.franky-f29.workers.dev · Repos: private (GitHub DS2036/r4m-engine — reviewer access on request)
Plak dit blok (of de links) in elk vrij tekstveld / "additional info". Alles hieronder komt live van de server — elke beoordelaar kan het in 60 seconden zelf nadoen.
1. A real agent hitting a real x402 paywall (one command):
curl -i -X POST https://r4m-franky.fly.dev/inbox/franky-kaart \
-H "Content-Type: application/json" -d '{"from":"reviewer","message":"hi"}'
Response — a live, spec-conformant HTTP 402 challenge (captured 2026-07-03):
HTTP/2 402
{"x402Version":1,"error":"X-PAYMENT header is required","accepts":[{
"scheme":"exact","network":"base-sepolia","maxAmountRequired":"2000000",
"resource":"http://r4m-franky.fly.dev/inbox/franky-kaart",
"description":"Inbox402: pay the recipient to deliver a message to their attention — the payment lands directly with them.",
"payTo":"0x34e3B98c3f621aF45786EC8C28B0EbB14aaD8B0c","maxTimeoutSeconds":60,
"asset":"0x036CbD53842c5426634e7929541eC2318f3dCF7e","extra":{"name":"USDC","version":"2"}}]}
2a. Real x402-settled data sales (agent-paid, July 6-7): an autonomous buyer paid $0.01 per query over HTTP 402 — settled on Base Sepolia, booked as x402_data_sale lines on the public ledger:
https://r4m-franky.fly.dev/api/public/ledger — and the live counter: https://r4m-franky.fly.dev/api/public/honest-math (x402DataSales).
2. A real paid transaction with the split applied end-to-end (85/8/4/3 at the time; 88/8/4 since 6 July), on the public ledger:
https://r4m-franky.fly.dev/api/public/ledger — first real card payment (2026-07-02): $2.00 in → $1.70 to the recipient (85%) · $0.16 creator fee (8%) · $0.08 state share (4%) · $0.06 birth vaults (3%). Every entry timestamped; card revenue ledgered separately from x402 revenue (never mixed).
3. Honesty as an API — measured reality, never projections:
curl https://r4m-franky.fly.dev/api/public/honest-math curl https://r4m-franky.fly.dev/api/public/solvency
The engine publishes its own truth: settled vs pending never added together; solvency shows committed obligations vs treasury — a deficit is displayed, never hidden. Current honest numbers: live at /api/public/honest-math (hand-copied snapshot removed 24-08-2026) — pre-revenue, and the API says so itself.
3b. BOTH Circle currencies proven on-chain (July 4): an automated USDC settlement (obligation-first settler; recipient was a test placeholder — first third-party payout is milestone M1's opening step, we never overstate) and an EURC transfer flowed through the SAME engine — switching currency is a single config change. The rail is globally pluggable: EURC for EU pilots today, regional stablecoins as next plugs. Every deployment is distribution for Circle's coins.
• USDC payout (auto-settled by the obligation-first settler): https://sepolia.basescan.org/tx/0x293391809abfb97a38bfdfa289c0d6867e6bcf451fd09057a0e8e79f5ace5c19
• EURC transfer (same engine, adapter switched): https://sepolia.basescan.org/tx/0x082692ddaa2e4c89813398859a4e3cd9288ba44127f71252f2e92332b3980599
4. Config transparency: https://r4m-franky.fly.dev/api/public/config — shows adapter (usdc-base), network, the split, and x402-priced data products (from $0.005 per call).
Planned legal gate (explicit): a Belgian fintech-counsel perimeter opinion (MiCA/PSD2) is the hard gate before any third party holds a balance on the rail — we rent licenses (Stripe Connect fiat, licensed facilitators for stablecoins), we never become the custodian. Founder self-tests on testnet involve no client funds.
What runs today vs. what's roadmap (no blur): today = engine v0.27 (162 automated tests), testnet USDC + card testmode, EURC adapter shipped, on-chain deposit verification live-tested on Base Sepolia, and itsme attestation flow proven end-to-end in the Signicat sandbox on 24-08-2026 (real QR scan with the itsme test app; attestation issued and publicly verified). Roadmap = mainnet micro-pilots, itsme production activation (broker onboarding in progress), Stripe Connect payouts. We never present roadmap as done. We never present roadmap as done.
Bron: pitch/templates/GRANT_CIRCLE_APPLICATION.md (privé-repo) — deze pagina wordt bij elke wijziging opnieuw gegenereerd via build-docs.py.