Does ai-job-search Have Deployment Scripts? Local Setup and CI Guide

The ai-job-search repository does not contain deployment scripts or cloud deployment instructions; it is designed to run locally as an interactive job-search assistant using Claude Code.

The ai-job-search project by MadsLorentzen is an open-source job search automation tool built to operate entirely on your local machine. Unlike traditional web applications, this repository lacks server-side entry points, Dockerfiles, or Kubernetes manifests, making ai-job-search deployment to remote servers impossible without custom modifications. Instead, the project relies on a local-first architecture centered around the Claude Code CLI and interactive setup workflows.

Local-First Architecture: Why Deployment Scripts Are Absent

The codebase is explicitly architected for local execution. In SETUP.md, the installation guide directs users through dependency installation—including Claude Code, Python, Bun, and LaTeX—without mentioning containerization or remote hosting. The repository root contains no Dockerfile, docker-compose.yml, or cloud-provider configuration files such as AWS CloudFormation or Terraform scripts. The application entry point requires an interactive Claude Code REPL session, which fundamentally conflicts with headless server deployment patterns.

Setting Up ai-job-search Locally

Since no ai-job-search deployment scripts exist, installation follows a manual local setup process. The primary onboarding documentation lives in SETUP.md at the repository root.

Prerequisites

Before cloning the repository, ensure your system has:

• Claude Code installed and authenticated
• Python 3.8+ with pip
• Bun runtime for TypeScript execution
• LaTeX distribution (for PDF generation features)

Installation Steps

Clone the repository and initiate the interactive setup:

# Clone the repository

git clone https://github.com/MadsLorentzen/ai-job-search.git
cd ai-job-search

# Launch Claude Code REPL

claude

# Run the interactive onboarding

/setup

The /setup command triggers an interview-style configuration process that customizes the tool for your specific job search parameters. This interactive requirement prevents traditional automation via deployment scripts.

CI Workflow Analysis: Testing Without Deployment

The repository includes continuous integration via .github/workflows/ci.yml, but this workflow strictly validates code quality rather than packaging artifacts for deployment.

What the CI Pipeline Does

The GitHub Actions configuration executes the test suite on every push to verify functionality. As implemented in MadsLorentzen/ai-job-search, the workflow runs pytest against the Python codebase and validates the TypeScript CLI utilities in .agents/skills/*/cli/. Crucially, the workflow lacks build stages for container images, deployment steps to cloud environments, or artifact publishing actions.

Running Tests Locally

To mirror the CI validation on your local machine:

# Install Python dependencies (if requirements.txt exists)

pip install -r requirements.txt

# Install Bun dependencies for TypeScript CLIs

bun install

# Execute the full test suite

pytest tests/

Project Structure and Entry Points

Understanding the directory layout clarifies why server deployment is impractical:

• SETUP.md – Contains the complete local installation guide
• .github/workflows/ci.yml – GitHub Actions configuration for testing only
• tools/ – Helper scripts for salary conversion and upstream monitoring
• .agents/skills/*/cli/ – TypeScript-based command-line interfaces for individual job portals

None of these directories contain WSGI/ASGI entry points, HTTP server configurations, or background worker implementations typically required for deployed applications.

Summary

• ai-job-search contains no deployment scripts, Dockerfiles, or Kubernetes manifests
• The application is architected exclusively for local execution via Claude Code interactive sessions
• Setup requires manual dependency installation guided by SETUP.md
• The .github/workflows/ci.yml file handles testing but does not package or deploy the application
• Remote execution requires repeating the local setup process on the target machine

Frequently Asked Questions

Can I deploy ai-job-search to a cloud server or VPS?

No. The tool requires an interactive Claude Code REPL session and local access to system dependencies like LaTeX and Bun. Without significant refactoring to remove interactive dependencies and create headless entry points, ai-job-search deployment to remote servers is not supported by the existing codebase.

Does ai-job-search support Docker or containerization?

No. The repository contains no Dockerfile, .dockerignore, or container orchestration files. The interactive nature of the /setup command and the dependency on local Claude Code authentication make containerization non-trivial for the current architecture.

How do I run ai-job-search on a different machine?

Since no deployment automation exists, you must manually replicate the local environment. Copy the repository to the new host, install the prerequisites listed in SETUP.md (Claude Code, Python, Bun, LaTeX), and run the /setup command within the Claude Code REPL to configure the instance.

What is the purpose of the CI workflow in ai-job-search?

The .github/workflows/ci.yml file validates code changes by running automated tests against Python and TypeScript components. According to the source code analysis, this workflow ensures code quality but explicitly excludes build, package, or deployment stages to remote environments.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}