gclaw-agent

Configure and run Gclaw agent DeFi trading with live-mode gating.

2|Updated Feb 25, 2026
One-click install
npx skills add https://github.com/terraleiloa/seren-skills --skill gclaw-agent
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: gclaw-agent
Source: https://github.com/terraleiloa/seren-skills/tree/main/gemachdao/gclaw-agent
Command: npx skills add https://github.com/terraleiloa/seren-skills --skill gclaw-agent

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

It solves the problem of running an autonomous trading-and-task agent continuously by handling setup, configuration, and safe execution for DeFi survival mechanics.

Core Features & Use Cases

  • Installation & setup guidance: Install the Gclaw Go binary and initialize the agent workspace and configuration.
  • LLM provider and trading configuration templates: Generate a working config.json via onboarding with model_list entries for providers and settings for DeFi trading through GDEX.
  • Safety-guarded execution and emergency controls: Enforce live-mode gating (config + explicit --yes-live) and provide unwind behavior to cancel orders and liquidate inventory.

Use case: you want an always-on agent that can run in dry-run mode for testing, then switch to live trading only after explicit approval, while optionally using cron, channels, and a local dashboard.

Quick Start

Start by running the agent onboarding wizard so it creates your ~/.gclaw/config.json for your chosen LLM provider.

Frequently Asked Questions about gclaw-agent

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I set up an autonomous DeFi trading agent for continuous operation?

To set up an autonomous DeFi trading agent, run the onboarding wizard to initialize the workspace, generate a config.json with your LLM provider settings, and validate required credentials before execution.

What safety mechanisms prevent an autonomous trading agent from executing unintended live transactions?

Autonomous trading agent safety relies on fail-closed behavior that enforces live-mode gating, requiring explicit opt-in via the execution.live_mode configuration setting and the --yes-live runtime flag to authorize transactions.

Can I test a DeFi trading strategy in dry-run mode before enabling live execution?

Yes, you can test DeFi trading strategies in dry-run mode for testing, allowing the agent to operate safely before explicitly switching to live trading only after granting manual approval.

How do I safely stop a long-running DeFi agent and liquidate its active positions?

To safely stop a long-running DeFi agent, use the emergency unwind behavior to cancel open orders and liquidate inventory, ensuring the autonomous trading workflow halts without leaving active positions.

Do I need specific LLM provider credentials to run an autonomous DeFi trading agent?

Yes, running an autonomous DeFi trading agent requires specific LLM provider credentials configured in the model_list entries within your config.json, which must be validated before trading execution begins.

Can I coordinate multiple autonomous trading agents in a swarm for decentralized finance operations?

Yes, you can coordinate multiple autonomous trading agent instances using optional swarm coordination, enabling multi-agent management alongside long-running gateway operation with cron and multi-channel communication.