self-hosted

Self-Hosted AI Agent Tools

Run the full Slopshop compute layer on your own servers. 925 pure compute APIs, zero external dependencies, complete data sovereignty. One Node.js process, one SQLite file, five minutes to deploy.

Why self-host?

Three reasons teams choose to self-host rather than use the cloud API:

What you get when self-hosting

The self-hosted deployment includes all 925 pure compute handlers -- text processing, math, code utilities, crypto, dates, data analysis, schema validation, PII detection, and more. These handlers have zero external dependencies. They do not call any third-party APIs or services.

You also get persistent memory (KV store, queues, counters) backed by the local SQLite database, audit logging, and the full REST API surface. The MCP server for Claude Code works identically against a self-hosted instance.

Deployment in 5 minutes

Deploy locallybash
# Clone and install
git clone https://github.com/slopshop/slopshop.git
cd slopshop
npm install

# Start the server
node server-v2.js

# Custom port
PORT=8080 node server-v2.js

# With AI APIs (optional)
ANTHROPIC_API_KEY=sk-... node server-v2.js

System requirements

Docker deployment

For production deployments, use the official Docker image. It bundles Node.js, the server, and all dependencies into a single container. Mount a volume for SQLite persistence and you are done.

Dockerbash
docker run -d \
  -p 3000:3000 \
  -v slopshop-data:/app/data \
  slopshop/slopshop:latest

Air-gapped environments

The compute layer works in fully air-gapped environments with no internet access. Since all 925 compute handlers run locally with no external API calls, there is nothing to connect to. This makes Slopshop uniquely suitable for classified environments, secure labs, and isolated networks where cloud APIs are not an option.

Network tools (DNS lookups, HTTP probes, SSL checks) obviously require network access, but they are a separate handler module you can disable entirely in air-gapped deployments.

Hybrid deployment

Many teams use a hybrid approach: self-host for compute-heavy workloads and data-sensitive operations, use the cloud API for network tools and features that require internet access. The API surface is identical, so switching between endpoints is a one-line config change. Get started with the cloud API and migrate to self-hosted when ready.

Start self-hosting today

Clone, install, run. 925 compute APIs in 5 minutes. Or start free on the cloud.

$ git clone https://github.com/slopshop/slopshop.git