Real payments for OpenClaw agents.
Prepaid wallet. Virtual card. OTP and 3DS inbox. Auditable payment flows.
Website · Docs Site · OpenClaw Quickstart · Docs · Skill
This is the public MoneyClaw repository.
It is the place for:
- public product docs
- the MoneyClaw OpenClaw skill
- examples and starter prompts
- high-level architecture and safety model
It is not the full internal product monorepo.
Internal infrastructure, production runbooks, incidents, security hardening notes, deployment details, and provider-specific operational playbooks are intentionally not published here.
MoneyClaw gives OpenClaw agents a real prepaid wallet and virtual card for online purchases.
With MoneyClaw, an agent can:
- check wallet and card readiness
- issue a virtual card
- top up a card with prepaid funds
- complete browser checkout
- fetch OTP or 3DS codes from a dedicated inbox
- verify the final transaction result
Secondary workflows include:
- hosted invoices
- merchant checkout links
- merchant and acquiring flows when explicitly requested
Install the skill:
clawhub install moneyclawThen:
- Create a MoneyClaw API key.
- Store it in OpenClaw as
MONEYCLAW_API_KEY. - Start a fresh OpenClaw session.
- Ask the agent to inspect the account before its first purchase.
Starter prompt:
Check my MoneyClaw account, show wallet balance vs card balance, and tell me if it is ready for a purchase.
MoneyClaw is built around bounded, inspectable agent payments.
- prepaid balances keep risk capped
- wallet and card history stay separate
- OTP and 3DS verification flows use a dedicated inbox
- the user can operate in approval-based or pre-authorized mode
Important boundary:
MoneyClaw is for user-authorized payments. It is not designed to bypass issuer, merchant, fraud, KYC, sanctions, geography, or verification controls.
Read the full model in docs/security-model.md.
Start here:
assets/ branding and repo visuals
docs/ public documentation
examples/ prompts and curl examples
moneyclaw-skill/ public OpenClaw skill files
Published here:
- what MoneyClaw is
- how to use it
- how the public skill is structured
- how the public API is meant to be understood
- how the safety and trust model works
Not published here:
- production IPs and infra
- internal incidents and postmortems
- operational security notes
- private merchant-specific tactics
- internal team process and private agent instructions
- Product site: moneyclaw.ai
- Docs hub: elvismusli.github.io/moneyclaw
- Security contact: SECURITY.md
Code and docs in this repository are available under the MIT License, unless noted otherwise.