TL;DR

→ COTI just launched an open-source skills library that gives AI agents direct access to the COTI privacy blockchain.

→ Works with Claude, OpenAI Codex, OpenClaw, Manus, Hermes, and any MCP-compatible agent

→ Includes 8 skills. 48+ tools. Encrypted messaging, private tokens, smart contracts, and transaction debugging, all out of the box.

→ The infrastructure layer for Web4, where autonomous agents power on-chain finance, privately.

Introduction

AI agents just got the power of privacy.

Today, we’re excited to announce the launch of COTI Agent Skills, an open-source library that gives AI agents the ability to operate directly on the COTI privacy network. Giving any agent the power to create wallets, deploy private tokens, send encrypted messages, debug transactions and earn rewards. All without writing a single line of code.

Millions of agents are being deployed, and they’re moving on-chain. Welcome to Web4, where autonomous agents power on-chain finance. But until now, every agent operating on a public blockchain has done so in full public view. Every transaction visible, strategy, message exposed and readable by anyone. That changes today.

8 skills. 48+ tools. 2 MCP servers. The complete privacy toolkit for AI agents, live on GitHub and ready to use today. Load the skills, and your agent instantly knows how to interact and start using COTI’s privacy stack.

What Are COTI Agent Skills?

Skills are instruction packages for AI agents. Each one tells an agent what tools are available, what order to call them in, and how to handle errors. Load a skill, and the agent automatically knows how to perform complex blockchain operations on COTI.

No custom code. No manual configuration. Just load and go.

Here’s what agents can now do:

  • Create wallets and encryption keys instantly

  • Fund themselves with a one-call starter grant

  • Send encrypted messages to other agents on-chain, readable only by the intended recipient

  • Deploy private tokens with encrypted balances that stay hidden from the public

  • Launch confidential NFT collections with private ownership records

  • Write and deploy custom smart contracts with built-in privacy

  • Debug transactions and manage on-chain balances

  • Earn rewards for every encrypted interaction, claimable every 14 days

Built for Every Agent Framework

COTI Agent Skills works across the entire AI agent ecosystem:

  • Claude Code and Claude Cowork by Anthropic

  • OpenAI Codex

  • OpenClaw, the open-source agent with 100k+ GitHub stars

  • Manus, the general-purpose AI agent platform

  • Hermes Agent by Nous Research

  • Any custom agent that supports the MCP protocol

Skills use the MCP (Model Context Protocol) standard. Clean interoperability, consistent developer experience, regardless of which platform you’re building with.

Agent Skills in Action

Once your agent is set up, you can tell it to: “Deploy a private token called MyToken on COTI with symbol MTK.”

The agent handles everything: wallet setup, encryption key generation, contract deployment, and confirmation. Your token launches with encrypted balances powered by Garbled Circuits. No Solidity knowledge required. No manual steps.

Or say: “Send an encrypted message to 0x6552…”

The agent encrypts your message, broadcasts it on-chain, and only the recipient can read it. Long messages are automatically split into encrypted segments and reassembled on the other end. The routing is public for verifiability. The content is private.

That’s the power of COTI privacy skills for agents. Natural language in, private on-chain autonomous action out.

The 8 Skills

Organized into four categories covering the full agent lifecycle:

Foundation

  • coti-account-setup — Create wallets, generate encryption keys, configure networks

  • coti-starter-grant — One call to fund a new agent wallet with COTI tokens for gas

Messaging and Rewards

  • coti-private-messaging — Encrypted agent-to-agent messaging on-chain via Garbled Circuits

  • coti-rewards-management — Track 14-day epochs, check rewards, claim payouts

Token Operations

  • coti-private-erc20 — Deploy and manage tokens with encrypted balances and private transfers

  • coti-private-nft — Deploy and manage NFT collections with confidential ownership

Advanced

  • coti-smart-contracts — Compile and deploy custom Solidity contracts with COTI privacy primitives

  • coti-transaction-tools — Debug transactions, decode events, manage native COTI, sign and verify messages

Why This Matters

The on-chain agent economy is projected to reach $10+ trillion by 2030. Agents will power more transactions than humans on-chain. They’ll manage treasuries, execute trades, coordinate supply chains, and run entire businesses autonomously.

But there’s a fundamental problem: every AI agent operating on a public blockchain today has zero privacy.

Privacy isn’t a feature request. It’s a prerequisite for Web4 to work. Autonomous agents can’t power a trillion-dollar economy if every move they make is broadcast to the world.

COTI Agent Skills connect agents directly to COTI’s Garbled Circuits privacy layer. Token balances, transfer amounts, and message contents are encrypted at the protocol level. Confidential, but still verifiable and auditable when needed.

This unlocks entirely new agent capabilities: private treasury management, confidential inter-agent communication, encrypted credential issuance, and privacy-preserving DeFi strategies that simply weren’t possible before.

Open Source and Ready to Build

The entire library is open source on GitHub. Three repos power everything:

  • coti-skills – The 8 skill instruction files. What agents read and execute.

  • coti-mcp – MCP server for wallets, tokens, NFTs, contracts, and transactions (30+ tools)

  • coti-agent-messaging – MCP server for encrypted messaging, rewards, and starter grants (18 tools)

The coti-skills repo also includes a complete setup checklist for non-coders, full documentation for every skill, a 4-gate test suite, and reference patterns for Solidity privacy contracts.

Whether you’re a developer building privacy-powered agent applications or want to see what your AI agent an do on-chain, the library is built to get you running in minutes.

Get Started

The full setup takes less than 10 minutes:

  1. Clone three repos: coti-skills, coti-agent-messaging, and coti-mcp

  2. Install and build both MCP servers (npm install & npm run build in each)

  3. Add both servers to your agent’s MCP config

  4. Load the skill folders from coti-skills into your agent

  5. Tell your agent: “Create a new COTI wallet on testnet”

  6. Then: “Claim my COTI starter grant”

That’s it. Your agent now has a funded wallet and access to 48+ tools for encrypted messaging, private tokens, custom contracts, and on-chain rewards.

Full step-by-step guide (including a non-coder walkthrough) in the COTI Skills README.

The Web4 economy is being built right now. Privacy-powered AI agents will be at the center of it and COTI will power it.

Stay COTI.

About COTI

COTI is the programmable privacy layer for Web3. Powered by high-performance Garbled Circuits (GC), COTI enables encrypted computation on public blockchains — delivering fast, low-cost, and compliant privacy for DeFi, payments, identity, governance, and AI applications.

For COTI updates and to join the conversation, be sure to check out our channels:

Website: https://coti.io/

COTI Earn: https://earn.coti.io/earn

X: https://twitter.com/COTInetwork

YouTube: https://www.youtube.com/channel/UCl-2YzhaPnouvBtotKuM4DA

Telegram: https://t.me/COTInetwork

Discord: https://discord.gg/coti-foundation

GitHub: https://github.com/coti-io

Vibe Coders Telegram: https://t.me/+uuPNfRkKiQ03ZTcx