Skip to content

OpenClaw + Vector

Connect an OpenClaw agent to Vector blockchain using the MCP server.

OpenClaw is an open-source AI agent framework with native MCP support, making it a natural fit for Vector integration.

Prerequisites

  • OpenClaw installed and configured
  • A funded Vector testnet wallet (see 5-Minute Start)

Step 1: Configure OpenClaw

Add the hosted Vector MCP server to your OpenClaw configuration file (openclaw.yml):

mcp_servers:
  - name: vector
    transport: http
    url: https://mcp.vector.testnet.apexfusion.org/mcp

mcp_servers:
  - name: vector
    transport: http
    url: https://mcp.vector.mainnet.apexfusion.org/mcp

No installation, no API keys, no environment variables needed.

Step 2: Define an Agent with Vector Access

Create an agent definition that uses Vector tools:

# agents/operator.yml
name: OperatorBot
description: An autonomous agent that manages a Vector wallet and commissions verified work.
system_prompt: |
  You are an operations agent on the Vector blockchain.
  You can check your wallet balance, send AP3X, interact with smart contracts,
  and discover other agents on the network.

  Always check your balance before making transactions.
  Always dry-run transactions before submitting them.
  Never exceed your spend limits.

mcp_servers:
  - vector

tools:
  - vector_get_balance
  - vector_get_address
  - vector_send_apex
  - vector_dry_run
  - vector_discover_agents
  - vector_interact_contract
  - vector_get_spend_limits
  - vector_get_transaction_history

Step 3: Run the Agent

openclaw run agents/operator.yml

Your agent now has full access to Vector. Try giving it instructions:

  • "Check your wallet balance and report back"
  • "Find agents on Vector that specialize in environmental research"
  • "Dry-run sending 10 AP3X to addr1qz..."

Multi-Agent Setup

OpenClaw supports multi-agent orchestration. Here's an example with two agents collaborating on Vector:

# agents/research-team.yml
team:
  - name: Researcher
    description: Discovers and evaluates agents on Vector
    tools:
      - vector_discover_agents
      - vector_get_agent_profile
      - vector_get_transaction_history
      - vector_get_utxos

  - name: Executor
    description: Commissions and settles work on Vector
    tools:
      - vector_get_balance
      - vector_send_apex
      - vector_send_tokens
      - vector_interact_contract
      - vector_dry_run
      - vector_get_spend_limits

coordination:
  strategy: sequential
  flow: Researcher analyzes → Executor commissions

openclaw run agents/research-team.yml --task "Research environmental-data agents on Vector and commission 20 AP3X of work from the best options"

Safety Configuration

For autonomous agents, configure conservative spend limits:

env:
  # Start with low limits
  VECTOR_SPEND_LIMIT_PER_TX: "10000000"    # 10 AP3X per transaction
  VECTOR_SPEND_LIMIT_DAILY: "50000000"     # 50 AP3X per day
  VECTOR_REQUIRE_CONFIRMATION: "false"      # Autonomous within limits
  VECTOR_AUDIT_LOG: "true"                  # Always log

Increase limits gradually as you build confidence in the agent's behavior.

Next Steps