[ Lending intelligence for agents ]

Less noise. More signal.

Agents pay per call for live multi-protocol lending risk — then move USDC only when the data says go.

The Graph × Hedera × Arc · Messari standardized subgraphs

Open demo See the thesis

01 — Problem

When agents fly blind.

Most agent stacks dump stale JSON, burn API-key seats, or skip freshness entirely. Capital moves on guesses.

Machine economies need metered truth, not subscriptions. We built MeteredSignal so agents can buy standardized Graph intelligence on Hedera rails and only unlock Arc spend when the signal is fresh and actionable.

Market

$35B+

Live Messari lending TVL across Aave, Compound & Spark in one query shape

The Graph

1.27T+

Queries already served to 75,000+ projects across 50+ networks

Settlement

~3s

Hedera x402 via Blocky402 — sub-cent pay-per-signal, no seats

02 — Signal

Chaos, then verdict.

One Messari schema across protocols. _meta freshness gates. A GO / NO_GO / UNAVAILABLE decision — not a raw dump.

lending-compare wallet-risk GO · NO_GO · UNAVAILABLE pay per call · HBAR

03 — Stack

And then, capital moves.

On GO, the Arc agent spends native USDC under policy limits. Graph evidence is load-bearing. Hedera settles the meter. Arc executes.

aave-v3-ethereum · $24.9B spark-lend · $7.2B compound-v3 · $1.9B aave-v3-arbitrum · $0.9B

04 — Use cases

Who buys the signal.

01 — Treasury

Treasury agents

Score borrow exposure before reallocating stables across Aave, Compound, and Spark.

02 — Desks

Lending desks

Compare protocol TVL / borrow across chains before quoting risk — one Messari query shape.

03 — Spenders

Autonomous spenders

Gate USDC nanopayments on freshness-backed GO signals instead of blind automation.