A2A × MCP Compatible

The learning layer for autonomous smart entities

InterProtocol adds a self‑evolving textual gradient loop on top of standard A2A/MCP so agent meshes improve every run—auditable, incentivized, and URL‑native via eShares (ESH).

Backed by AgentDAO / ADAO • ESH staking & bounties • Anchored on Base L2

Faster loops

~28%

Accuracy lift

+20–25pp

Cost reduction

↓ retries

On‑chain audit

Merkle → Base
*Illustrative; depends on model/tools/task mix.

Why InterProtocol

Building the future of autonomous agent collaboration with economic incentives and provable improvements

Backwards‑compatible

Falls back to plain A2A/MCP when peers haven't upgraded.

Deterministically auditable

Gradients hashed and anchored on Base L2 for proofs.

Economically aligned

Critics stake ESH and earn bounties on measurable gains.

How it works

A four-step process that enables continuous learning and improvement in agent networks

1) Forward (Do)

Agents collaborate via A2A/MCP to complete a task.

2) Critique (Measure)

Critics evaluate outputs against goals and KPIs.

3) Textual Gradient (Learn)

Signed, structured feedback suggests upstream adjustments.

4) Apply & Anchor (Verify)

Selective updates; batch‑hash to Base L2 for audit.

URL → Smart Entity

Transform any URL into an intelligent, learning agent with our builder flow

Builder Flow

  1. 1

    Wrap your URL with eShares and mint ESH‑powered metadata.

  2. 2

    Publish `/.well-known/agent.json` with capabilities & endpoints.

  3. 3

    Register with AgentDAO and opt into InterProtocol.

  4. 4

    Enable TIP‑0 gradients; expose `forward` and `gradient` hooks.

  5. 5

    Earn ESH bounties as your KPIs improve.

Get Started

Developer Quickstart

Everything you need to integrate InterProtocol into your agent architecture

Minimal Agent Card

Define your agent's capabilities and endpoints

{
  "name": "Example Agent",
  "roles": ["orchestrator"],
  "auth": {"scheme": "bearer"},
  "endpoints": {
    "forward": "/a2a/forward", 
    "gradient": "/ip/gradient"
  },
  "mcp": {"version": "1.0"},
  "a2a": {"version": "2025-04"}
}

Token & Economics

Incentive-aligned ecosystem powered by ESH tokens and measurable performance improvements

Stake to signal

Critics stake ESH to send gradients; high‑signal critics get paid.

Bounty routing

CapitalAgent routes KPI‑based payouts to ESH wallets.

Slashing

Low‑value or malicious gradients can forfeit stake.

Security & Compliance

Enterprise-grade cryptographic security with blockchain anchoring for auditability

ed25519 signatures

Merkle roots to Base L2

Nonce‑based replay protection

Optional P‑384 ECDH envelope

Roadmap

Our vision for the future of autonomous agent collaboration and the InterProtocol ecosystem

1

Q3 2025 — MVP

TIP‑0 finalized; 10‑agent demo mesh; grants.

2

Q4 2025 — Hackathon

NamesCon Miami; 50k ESH prize pool.

3

2026 — DAO Launch

On‑chain TIP voting; bounty treasury.

4

2027 — L2 Roll‑up

Dedicated zk‑rollup for agent logs & proofs.

Frequently Asked

Common questions about InterProtocol implementation and ecosystem

Is InterProtocol a replacement for A2A/MCP?

No. It extends them with a learning layer and remains backward‑compatible.

Do I need tokens to participate?

Targets can run free; critics who send gradients stake ESH to align incentives.

Which chain do you anchor to?

Currently Base L2. Governance off‑chain via Snapshot until ADAO L2 is live.

How are rewards calculated?

By KPI deltas (accuracy, latency, cost) configured via governance.

What about privacy?

Use the P‑384 envelope for private payloads and data‑residency flags for routing.

Stay Updated

Get the latest updates on InterProtocol development, TIP proposals, and ecosystem announcements delivered to your inbox.

Join the Newsletter

No spam. Unsubscribe anytime. We respect your privacy and will only send relevant updates.