Adds a security layer for AI agents making crypto payments through ClawVault's API. Claude can use request_payment to send USDC or other tokens, check_limits to validate amounts before sending, and get_payment_status to track approvals. Payments run through your configured spending limits and whitelists. If a payment falls within your rules, it auto-approves. If not, you get a Telegram notification and approve or deny manually. Also exposes get_vault for balance checks, list_transactions for history, and freeze_agent as an emergency kill switch. Reach for this when you want Claude to handle payments autonomously but need guardrails to prevent runaway spending or unauthorized recipients.
MCP (Model Context Protocol) server for ClawVault - AI agent payment security layer.
npm install -g clawvault-mcp-server
Or use directly with npx (no install needed):
npx clawvault-mcp-server
Add to your Claude Desktop config:
Mac: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"clawvault": {
"command": "npx",
"args": ["clawvault-mcp-server"],
"env": {
"CLAWVAULT_API_KEY": "cv_live_your_api_key_here"
}
}
}
}
Restart Claude Desktop after saving.
cv_live_...)Once connected, Claude has access to these tools:
Request a payment through ClawVault. The payment will be evaluated against your security rules.
Amount: "10.00"
Token: "USDC"
Recipient: "0x..."
Chain: "base"
Reason: "API credits" (optional)
Check if a payment would be allowed before making it. Returns whether it would auto-approve or need manual approval.
Check the status of a pending payment by ID.
List recent payment transactions with pagination.
Get vault status including wallet address, balances, and current spending limits.
Emergency freeze an agent to block all its payment requests.
request_payment tool| Variable | Required | Default | Description |
|---|---|---|---|
CLAWVAULT_API_KEY | Yes | - | Your ClawVault API key |
CLAWVAULT_API_URL | No | https://api.clawvault.cc | API base URL |
Not using Claude Desktop? Any AI agent can use ClawVault by reading the skill documentation:
https://api.clawvault.cc/skill
This returns plain text that teaches the agent how to use the ClawVault API.
# Clone the repo
git clone https://github.com/clawvault/mcp-server
# Install dependencies
npm install
# Build
npm run build
# Run locally
CLAWVAULT_API_KEY=cv_live_xxx npm start
MIT
CLAWVAULT_API_KEY*secretYour ClawVault API key from the dashboard
io.github.ericm1018/skillfm-llm-cost-optimizer-openai-anthropic-usage
io.github.mikerawsonnz/llm-orchestration-agent
io.github.mikerawsonnz/authenticated-llm-agent
labforgedev/copilot-memory-mcp
csoai-org/agent-prompt-injection-firewall-mcp
io.github.mikerawsonnz/authenticated-multi-llm-agent