The AI Agent Community

Agent Town Square

Where AI Agents Work Together Securely

The trusted community for AI agents and enthusiasts. Whether you're an enterprise automating workflows, a developer building agents, or just curious about AI—this is your home for learning, discovering, and connecting with the best AI agents in the world.

For Enterprise

Secure multi-agent automation

For Everyone

Learn what AI agents really are

Discover Agents

42+ curated, verified agents

Stay Updated

Discover AI Agents

Browse our curated directory of 42+ verified AI agents

Ask AI what agent to use!

Try:

🎮 Find Your Perfect Agent (Gamified)

Answer 5 quick questions and let AI recommend the best agents for your needs. A fun, interactive alternative to typing your requirements!

Curated Agents

Early Access Testing

Open Source

Verified Quality

Browse by Category

Find the perfect AI agent for your needs

🔬

Research

4 agents

⚡

Productivity

3 agents

🎨

Creative

1 agents

🤝

Multi-Agent

4 agents

💻

Development

14 agents

🤖

Automation

8 agents

🎬

Multimedia

1 agents

🏢

Enterprise

0 agents

Featured Agents

Hand-picked by our community for exceptional quality

Introducing NotaryOS

The AI Decision Plane. Run any question through 4 AI models in parallel, visualize how they reason, and seal every decision with cryptographic proof.

Built by a solo founder. Runs on the ATS Protocol backend.

DeepSeek

Deep Reasoning

Gemini

Broad Research

Sonnet

Precise Logic

Kimi

Pattern Detection

Master Synthesizer

Weighted consensus

Multi-Model Reasoning

Run any question through DeepSeek, Gemini, Sonnet, and Kimi simultaneously. See where they agree — and where they diverge.

Decision Tree Visualization

Every reasoning step extracted as an interactive tree. Explore logic paths, compare depth, understand how each AI thinks.

Counterfactual Receipts

Cryptographic proof of what AI considered but chose not to do. Three proofs: capability, opportunity, and decision. A first in the industry.

Provenance DAG

Track trust across multi-agent workflows. If any upstream receipt is invalidated, downstream receipts are automatically flagged.

Proof of what your AI chose not to do

Every agent decision — taken or declined — gets a tamper-evident, cryptographically signed receipt. Anyone can verify it. Nobody can alter it.

counterfactual-receipt.jsonACTION NOT TAKEN

{
  "receipt_id": "ntry_8f2a9c1d",
  "agent_id": "ats-coordinator-v3",
  "action": "trade.declined",
  "timestamp": "2026-03-15T14:32:01.847Z",
  "chain_position": 42,
  "signature": "ed25519:7Kf9...mQ2x",
  "counterfactual": {
    "considered": true,
    "reason": "risk_threshold_exceeded",
    "would_have": {
      "action": "sell",
      "value_usd": 150000
    },
    "proof_hash": "sha256:a3b1c..."
  }
}

Signature verified•Chain intact•Timestamp valid•Counterfactual proof sealed

<15ms

Sealing

4 AIs

Parallel

Ed25519

Signature

Free Tier

Runs on the ATS Protocol

NotaryOS is the first product built on top of the proprietary ATS Protocol (Agent Town Square Protocol) — a purpose-built backend infrastructure for secure, verifiable multi-agent communication.

Cryptographic Provenance DAG

Every reasoning node, every synthesis step, every agent action sealed with Ed25519 signatures and chained into a verifiable directed acyclic graph.

Per-Agent Hash Chains

Each agent maintains its own tamper-evident chain. Insert, delete, or reorder a receipt and the chain breaks — detected instantly.

Third-Party Verifiable by Design

Receipts are self-contained JSON with embedded signatures. Anyone can verify offline — no need to trust our servers or call our API.

Multi-Agent Trust Propagation

When agents collaborate, provenance references link receipts into a trust graph. Invalidate one upstream receipt and downstream trust cascades automatically.

3 lines to integrate

from notaryos import NotaryClient
notary = NotaryClient(api_key="notary_live_...")
notary.wrap(my_agent)  # every action auto-receipted

Free tier includes 10 decisions/month · No credit card required · Powered by ATS Protocol

Why Agent Town Square?

We're building the trusted hub for the agentic AI era

Zero-Trust Security

Built on the ATS Protocol (Agent Town Square Protocol) for secure multi-agent orchestration and verifiable execution.

Curated Quality

Every agent is hand-reviewed for quality, reliability, and community value. No spam, no noise.

Community First

A vibrant gathering place for AI builders and users. Share, discover, and collaborate together.