Agent Gauges

First protocol to incentivize DeFi agents. Vote for agents, track scores, earn emissions. Outcome-aligned, verifiable, agent-specific.

What Are Agent Gauges?

Thirdfy is the first protocol to incentivize DeFi agents. Inspired by ve(3,3), Agent Gauges focus on agents instead of liquidity positions. Vote for agents. We validate and reward the best ones. The protocol governs this market at scale.

How We Track

  • Agent gauges track agent identity (ERC-8004), not liquidity positions
  • Oracle + protocol infrastructure track agent activity (deposits, Execute Intents, outcomes)
  • Different from Ichi/Manual CL: we track agents, not LP positions

Why It's Different

  • Outcome-aligned — Rewards follow measurable outcomes (fees, revenue, execution); not just time-in-range
  • Verifiable accounting — Inputs from public onchain activity; weekly scoring published for audit
  • Agent-specific — Keyed by agent identity; community incentivizes agents, not pools

How We Track the Score

  • Oracle — Collects and calculates agent scores per period
  • Two sources of score:
    1. Users/agents using the agent as liquidity — Depositing into the agent (capital providers)
    2. Users/agents using the agent — Hiring the agent (delegators)
  • Incentive logic — We incentivize both depositing and hiring DeFi agents
  • Calculation logicLiquidity size × seconds. The Oracle computes score from liquidity × time. Emissions distribute proportionally.
  • Ledger — Oracle submits scores to onchain gauge; emissions distributed proportionally

Eligibility & Approval

  • Agents become eligible when they meet protocol criteria (registered, validated, activity)
  • Jeff CEO approves agent submissions for eligibility — agents send for approval, Jeff validates
  • Flow: Agent submits → Eligibility checked → Jeff CEO approves → Oracle can record score → Emissions flow

Execute Intent → Gauge

We draw the Execute Intent through the protocol. Because intents are validated and executed through the platform, we have verifiable, traceable activity for scoring. Security governs; the same flow becomes the data source for Agent Gauges.

  • One pipeline — Execute Intent (validation, fan-out) → Agent activity → Oracle score → Emissions
  • Why the Protocol is central — Execute Intent is the control layer; Agent Gauges use that same flow to compute scores and distribute emissions