Skip to content

Latest commit

 

History

1 Commit

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 

Repository files navigation

hiivmind-bots-mulerun

LLM-powered assistant agent for MuleRun, built using the Agent Builder approach.

What This Is

A MuleRun agent that acts as a general-purpose AI assistant with modular skills. The agent runs in MuleRun's managed sandbox and is configured through natural language instructions + skill packages.

Repository Structure

hiivmind-bots-mulerun/
├── agents/
│   └── assistant/
│       └── AGENTS.md           # Agent instructions (role, behavior, constraints)
├── skills/
│   └── cowsay/                 # Cowsay ASCII art skill
│       ├── SKILL.md            # Skill instructions
│       └── scripts/            # Executable scripts
└── scripts/
    └── package-skill.sh        # Package skills into ZIP for upload

How Agent Builder Works

  1. Agent Instructions (AGENTS.md) - Define role, behavior, skill usage, and constraints
  2. Skills - Modular capability packages (ZIP with SKILL.md + optional scripts/assets/references)
  3. MCP Servers - External tool integrations (optional)
  4. Environment Variables - API keys and secrets (optional)

MuleRouter API

Agents have access to the MuleRouter API at https://api.mulerun.com:

  • LLM: OpenAI-compatible (/v1/chat/completions) and Anthropic-compatible (/v1/messages)
  • Image gen: Midjourney, Nano Banana, Wan 2
  • Video gen: Kling, Sora, Veo 3, Wan 2
  • MCP: Firecrawl (web scraping), Context7 (research)
  • Async tasks: Polling-based (pending → processing → completed/failed)

Deployment

Prerequisites

  • MuleRun Creator account (register here)
  • API key from Creator Studio

Steps

  1. Register as a Creator at https://mulerun.com/creator
  2. Generate API key in Creator Studio under Settings
  3. Create agent using Agent Builder UI
  4. Copy agent instructions from agents/assistant/AGENTS.md into the Agent Instructions field
  5. Package skills:
    ./scripts/package-skill.sh cowsay
  6. Upload skill ZIPs in the Skills section of Agent Builder
  7. Configure environment variables (set MULEROUTER_API_KEY to your API key)
  8. Build the agent and wait for the build to complete
  9. Test in the MuleRun chat interface
  10. Submit for review when ready to publish

Environment Variables

Variable Description
MULEROUTER_API_KEY MuleRouter API key for LLM/media generation calls

Adding New Skills

  1. Create a new directory under skills/:
    skills/my-skill/
    ├── SKILL.md
    ├── scripts/
    ├── references/
    └── assets/
    
  2. Write the SKILL.md with instructions for the agent
  3. Add any scripts, references, or assets
  4. Package: ./scripts/package-skill.sh my-skill
  5. Upload the ZIP in Agent Builder

Documentation

About

LLM-powered assistant agent for MuleRun with cowsay skill

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages