Deploy Agent to Runloop

A GitHub Action to deploy agents to the Runloop platform. Supports multiple source types including Git repositories and file uploads.

Features

• Zero-config Git deployments - Automatically deploys your current repository
• Release tag support - Deploys specific versions when releases are published
• Multiple source types - Git repositories, tar archives (.tar, .tar.gz, .tgz), and single files
• Flexible packaging - Create tar archives however you want in your workflow
• Setup commands - Run custom setup commands after agent installation
• Public/private agents - Control agent visibility
• TTL support - Set expiration time for uploaded objects

Quick Start

Deploy Current Repository as an Agent

- name: Deploy agent
  uses: runloopai/deploy-agent@main
  with:
    api-key: ${{ secrets.RUNLOOP_API_KEY }}
    source-type: git
    agent-version: 1.0.0

That's it! The action will automatically use your current repository and commit SHA.

Usage

Inputs

Input	Required	Default	Description
api-key	✅		Runloop API key (store in secrets)
source-type	✅		Agent source type: git, tar, or file
agent-version	✅		Agent version (semver string like 2.0.65 or git SHA)
agent-name		repo name	Name for the agent (defaults to repository name)
git-repository		current repo	Git repository URL (auto-detected)
git-ref		current commit/tag	Git ref (branch/tag/commit SHA, auto-detected)
path			Path to tar archive or single file (required for tar/file)
setup-commands			Newline-separated setup commands to run after installation
is-public		false	Whether the agent should be publicly accessible
api-url		https://api.runloop.ai	Runloop API URL
object-ttl-days			Time-to-live for uploaded objects in days

Outputs

Output	Description
agent-id	The ID of the created agent (e.g., agt_xxxx)
agent-name	The name of the created agent
object-id	The ID of the uploaded object (if applicable, e.g., obj_xxxx)

Examples

💡 See the examples/ directory for complete, ready-to-use workflow files!

Git Source (Auto-detect)

- uses: actions/checkout@v4
- uses: runloopai/deploy-agent@main
  with:
    api-key: ${{ secrets.RUNLOOP_API_KEY }}
    source-type: git
    agent-version: 1.0.0
    setup-commands: |
      chmod +x scripts/agent.sh
      npm install

Git Source (On Release)

on:
  release:
    types: [published]

jobs:
  deploy:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: runloopai/deploy-agent@main
        with:
          api-key: ${{ secrets.RUNLOOP_API_KEY }}
          source-type: git
          agent-version: ${{ github.event.release.tag_name }}
          agent-name: my-agent-${{ github.event.release.tag_name }}

Tar Archive (with custom packaging)

📋 See examples/tar-agent.yml for complete workflow examples with tar creation!

Basic example:

# Create your own tar.gz archive first
- name: Create agent archive
  run: |
    tar -czf agent.tar.gz -C ./agent-code .

# Then deploy it
- name: Deploy agent
  uses: runloopai/deploy-agent@main
  with:
    api-key: ${{ secrets.RUNLOOP_API_KEY }}
    source-type: tar
    agent-version: 1.0.0
    path: agent.tar.gz
    object-ttl-days: 30

You can also use .tar format or reference output from a previous step:

- name: Build and package
  id: build
  run: |
    # Your custom build process
    make package
    echo "archive-path=dist/my-agent.tar.gz" >> $GITHUB_OUTPUT

- uses: runloopai/deploy-agent@main
  with:
    api-key: ${{ secrets.RUNLOOP_API_KEY }}
    source-type: tar
    agent-version: 1.0.0
    path: ${{ steps.build.outputs.archive-path }}

Single File

- uses: runloopai/deploy-agent@main
  with:
    api-key: ${{ secrets.RUNLOOP_API_KEY }}
    source-type: file
    agent-version: 1.0.0
    path: ./scripts/agent.sh

Authentication

Store your Runloop API key as a GitHub secret:

1. Go to Settings → Secrets and variables → Actions
2. Create a new secret named RUNLOOP_API_KEY
3. Paste your Runloop API key
4. Reference it: api-key: ${{ secrets.RUNLOOP_API_KEY }}

Documentation

• 💡 Example Workflows
• 🌐 Runloop Docs

Development

Setup

pnpm install

Testing with act

You can test the action locally using act:

act -j deploy --secret RUNLOOP_API_KEY=your_api_key

Building

The action uses @vercel/ncc to bundle all dependencies into a single file.

pnpm run build        # Bundle with ncc (creates dist/index.js)
pnpm run rebuild      # Clean and build

After building, commit the dist/ folder as it's required for the action to run.

Code Quality

pnpm run lint         # Check for lint issues
pnpm run lint:fix     # Auto-fix lint issues
pnpm run format       # Format code with Prettier
pnpm run format:check # Check formatting
pnpm run typecheck    # Run TypeScript type checking
pnpm run check        # Run all checks (format + lint + typecheck)

License

MIT