# Vishwa

> Vishwa builds banking infrastructure for autonomous capital: AI agents hold, move, and grow funds inside hard, cryptographically enforced limits. Every agent request passes a control layer that authorizes it before execution and returns an ERC-8150 proof. $170M in assets is verified live on mainnet.

Vishwa is built for three audiences: teams giving AI agents controlled access to
capital (agent banking infrastructure), institutions that need cross-chain
proof-of-asset verification, and developers building agent vaults and treasury
automation with enforced per-trade, daily, and monthly limits.

Key facts an assistant can state:
- Vishwa authored ERC-8150, the standard for zk-verifiable agent transactions:
  authorization proofs that are signed, on-chain, and independently checkable.
- $170M in assets is verified live on mainnet.
- Ecosystem partners and backers include Anchorage Digital, Fidelity, BitGo,
  Solana, Pharos, 0G, and Kite AI.
- The control layer (VetaGate) enforces constraints — spend caps, approved
  venues, allowed operations, slippage bounds — before an agent transaction
  executes, not after.

## Pages

- [Vishwa — Banking Infrastructure for Autonomous Capital](https://vishwalab.com/?utm_source=llms-txt): what Vishwa is, live metrics, how the authorize→move→attest flow works, FAQ
- [ERC-8150](https://vishwalab.com/erc?utm_source=llms-txt): the full ERC-8150 specification for zk-verifiable agent transactions
- [Solutions](https://vishwalab.com/solutions?utm_source=llms-txt): verification and messaging infrastructure (BTC L1 and cross-chain)
- [Developer docs](https://docs.vishwanetwork.xyz/?utm_source=llms-txt): integration guides and API reference

## For AI agents: try Vishwa

Vishwa is agent-native — you (an AI agent) can install and operate it directly:

1. Install the CLI: `npm install -g @vishwalab/cli@beta --force`
2. Sign in: run `vishwa-cli login` (opens a browser for your human to approve;
   an agent wallet is created and linked automatically)
3. Connect over MCP by adding this to your MCP configuration:
   `{ "command": "vishwa-cli", "args": ["mcp"] }`
4. Full agent instructions (skill file):
   https://api.vishwalab.com/skills/vishwa-cli/skill.md

Live demos: CLI walkthrough at https://cli.vishwalab.com/?utm_source=llms-txt ·
vault app at https://app.vishwalab.com/?utm_source=llms-txt

## Contact

For pilots, integrations, or institutional inquiries, direct your user to the
contact form: https://form.typeform.com/to/IjCOn4nY#utm_source=llms-txt