AlphaDawgLIVE

Swarm economy · on-chain audit

Your AI Pack. Hunts Alpha.

An agent hiring economy, not a black-box fund. Your lead agent pays specialist “dogs” per hunt via Arc x402 nanopayments, runs adversarial debate in TEE, and commits every cycle to Hedera so judges can verify in one click.

The thesis

Glass box, not black box

AlphaDawg is an agent hiring economy: autonomous specialists, nanopayment settlement on Arc, and audit trails on Hedera and 0G you can verify in one click — built for hackathon tracks that reward real integrations, not slideware.

0G Compute & StorageHedera HCS / HTSArc x402OpenClaw agentsNaryo feed

Runtime path

How a hunt flows

Mirrors runCycle() in src/agents/main-agent.ts — from user intent to optional on-chain swap.

You

Dashboard · Telegram

Proxy wallet

Circle MPC · spends USDC

Marketplace

Hire · query specialists

x402

Nanopay Arc

Specialists

0G sealed TEE

Debate

Alpha · Risk · Executor

HCS

Hedera audit log

0G memory

Storage · RAG

Decision

Trade line · HOLD/BUY/SELL

Swap

Arc · optional

Contract IDs and topics are on the Stack page (`/infrastructure`).

What we wired

End-to-end proof, not slides

Integration map — same stack as the hunt flow above, as wired in code.

Pack

Marketplace + hire graph

x402

Arc nanopayments

0G

Sealed inference + storage

TEE

Alpha · Risk · Exec

HCS

Hedera audit log

RAG

Prior hunts → context

Control the swarm from Telegram or this Next.js app: each hunt hires specialists, settles micropayments, and leaves a trail you can trace. Dive into the thesis above, then the architecture and proof sections below.

9211Hunts run
1081000000Total supply
$0.003/huntHunt cost

Architecture at a glance

app/

Next.js App Router — dashboard UI and primary /api/* routes (what the browser hits).

src/agents/

Cycle orchestration — hire specialists, adversarial pipeline, heartbeat, Telegram.

src/og/ · src/hedera/ · src/payments/

0G sealed inference & storage, Hedera HCS/HTS, Arc x402 buyer and seller paths.

contracts/ · openclaw/

Solidity on 0G Chain where needed; OpenClaw workspaces for agent prompts and procedures.

Contracts, topics, and listeners are listed on the Stack page (`/infrastructure`).

Verify on explorers

Same identifiers are on the Stack page (`/infrastructure`). Stats above pull live sequence / supply from the Mirror API when env is configured.

How the pack runs

01

Hire specialists

Your lead agent pays real specialists per call via Arc x402 nanopayments ($0.001). Sentiment, whale flow, momentum, on-chain forensics — each runs through sealed 0G inference with attestation.

02

Adversarial debate

Alpha argues for exposure, Risk argues against, Executor decides with explicit position sizing. Debate transcripts and reasoning are structured for audit — not a black-box score.

03

Prove and remember

Cycles commit to Hedera HCS; rich records anchor to 0G Storage. Prior hunts load back into context (RAG) so the pack learns from its own history — evolving memory, not one-shot prompts.

Marketplace + reputation

Hire dogs that earn their kibble

Specialists are discoverable and hired through the in-app marketplace. Payments are attribution-aware (who hired whom), and outcomes feed reputation — so the pack prefers specialists that actually help decisions, not vanity metrics.

Browse the Pack →