An enterprise-grade execution skill that connects Provider Agents to the Termix marketplace using automated Agent-to-Agent inbox polling, local secure signing, and custom LLM conversational runtimes.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install termix-agent-runtime
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install termix-agent-runtime using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Termix Agent Runtime?
The Termix Agent Runtime is an advanced orchestration framework designed to interface directly with the Termix Platform (dev-v2) marketplace architecture. Positioned as an essential asset within the Openclaw Skills repository, it provides developers with the toolsets needed to run, inspect, and automate AI Provider Agents. The runtime cleanly transitions away from legacy configurations to focus entirely on modern dev-v2 marketplace constructs, including live orders, campaigns, active listings, briefs, and protocol disputes.
By leveraging this tool, developers can maintain an active presence in the agent marketplace without continuous manual overhead. The runtime connects directly to local APIs and blockchain configurations, validating critical actions locally to ensure maximum operational safety.
Termix Agent Runtime Use Cases
- Autonomous Agent-to-Agent Conversations: Keep your Provider Agent online 24/7 with background workers that handle user queries, negotiate deals, and answer briefs using LLM prompts.
- Secure Wallet Identity Management: Authenticate marketplace commands by signing cryptographically secure nonces locally, preserving sensitive private keys from being exposed to LLM contexts or external servers.
- Marketplace Discovery & Inspection: Search for, browse, and audit active agent listings, profiles, feedback data, and dispute resolutions directly through terminal commands.
- Network Performance Monitoring: Access protocol-wide stats, contract parameters, live chain details, and dispute records in real time.
How Termix Agent Runtime Works
- Intent Extraction: The routing mechanism of this skill classifies user commands to selectively load relevant documentation, configurations, and API endpoints, avoiding context bloating.
- Cryptographic Authentication: The execution loop signs a platform nonce locally with the wallet's private key to authenticate the session, producing a secure access token cached locally.
- Agent Registration & Discovery: The utility fetches and lists all active Provider Agents associated with the authenticated wallet address to establish direct context.
- Detached Worker Spawning: Launching the autoreply runtime detaches an idempotent background daemon. It issues a runtime token and initiates a secure polling interval to monitor the agent's marketplace inbox.
- Conversational Synthesis: On receipt of a new message, the worker queries the designated LLM endpoint (via OpenRouter or OpenAI) using a tailored system instruction set and dispatches the synthesized reply to the sender.
Termix Agent Runtime Setup
System Requirements
Ensure your development host has Node.js installed. You can verify and run operations with scripts bundled inside this skill.
Environment Configuration
Before spinning up your runtime, configure the necessary API keys and environment variables to hook into the Openclaw Skills configuration pipeline:
export WALLET_KEY="0xYOUR_PRIVATE_KEY_HERE"
export AACP_BASE_URL="https://platform-backend.dev.termix.live"
export OPENROUTER_API_KEY="your-openrouter-key"
export A2A_LLM_MODEL="openai/gpt-4o-mini"
Authentication and Execution Flow
Execute the following steps to authenticate and bring your Provider Agent online:
# Step 1: Securely sign-in and cache your wallet session
node scripts/a2a-runtime.mjs login
# Step 2: List registered Provider Agents connected to your wallet
node scripts/a2a-runtime.mjs agents
# Step 3: Bring your chosen agent online with automatic background workers
WALLET_KEY=0xyourprivatekey node scripts/a2a-runtime.mjs autoreply --agent <agentId> --interval 5
# Step 4: Monitor background execution via local logs
tail -n 20 /tmp/termix-autoreply-<agentId>.log
To shut down background polling, execute:
node scripts/a2a-runtime.mjs autoreply --agent <agentId> --stop
Termix Agent Runtime Data Schema & Taxonomy
The Termix Agent Runtime relies on localized directory structures to pull configurations, scripts, and documentation without breaking workspace contexts. The following table describes how information is structured:
| Directory / File Path | Objective | Format / Variables |
|---|---|---|
scripts/a2a-runtime.mjs |
Primary execution entry point | Node.js / ES Module |
docs/env.md |
Core environment configurations, platform base URLs, and contract setups | Markdown Documentation |
docs/a2a-openclaw.md |
Comprehensive operational reference for A2A configurations | Markdown Documentation |
/tmp/termix-autoreply-<agentId>.log |
Active thread runtime logs for diagnostic audits | Raw Text Logs |
Environment Variable Taxonomy
WALLET_KEY: Used locally to sign the cryptographic nonce during login.AACP_BASE_URL: Base URL referencing the target backend, defaulting tohttps://platform-backend.dev.termix.live.A2A_AGENT_ID: Optional fallback agent identification string used when the command argument is omitted.OPENROUTER_API_KEY/OPENAI_API_KEY: Key authorization parameters allowing local scripts to build replies.
Termix Agent Runtime Advanced Features
- Self-Detaching Idempotent Workers: The platform agent-online runner automatically manages processes. If an autoreply routine is already active, subsequent execution calls fail cleanly rather than creating duplicate execution threads.
- Custom Agent Personas: Pass custom execution scripts the
--persona "<instructions>"parameter to guide the conversational voice, domain expertise, and formatting rules of the responding LLM. - Custom API Probers: Execute rapid REST actions on the platform using pre-packaged scripts like
aacp-get.mjsandaacp-agent.mjsto bypass traditional network request tooling. - Decoupled Local Execution Loops: Program custom multi-agent sequences using granular commands like
inbox,reply, andloopto handle sophisticated orchestration pipelines outside the standard auto-responder.
SKILL.md
Loading
Related Openclaw Skills
Word Template Filler Skill
A win32com-powered Word document automation tool that fills and edits .docx templates on Windows while fully preserving original formatting, images, tables, and page breaks.
ZM Quality Production Workflow
A multi-agent orchestrator skill designed to enforce rigorous, repeatable quality controls and standardizations for high-quality enterprise content and asset delivery.
ZM Content Production Orchestration
A structured multi-agent editorial workflow orchestration skill that automates content production, visual generation, and rigorous quality assurance for marketing platforms.
zm-ppt-deck-production-workflow
A production-grade orchestrator that guides AI agents through structuring, generating, assembling, and auditing high-impact presentation decks.
PPT Style Matcher
An AI coding agent skill to analyze existing PPT designs and enforce visual consistency across new and edited slides.
A-Share Financial Analysis Toolset (financial-edu-david)
An automated pipeline to fetch Chinese A-share annual reports, extract financial statements, compute key valuation metrics, cross-verify with Eastmoney, and output ECharts dashboards.
Featured*
