Custom Endpoints - GitWhisper

GitWhisper allows you to configure custom API endpoints, making it possible to use self-hosted models, proxy servers, or alternative API implementations.

Overview

Custom endpoints are useful for:

Self-Hosted Ollama

Run Ollama on custom ports or remote servers

Corporate Proxies

Route API calls through company infrastructure

API Gateways

Use centralized API management systems

Development Testing

Test against local mock servers or staging environments

Ollama Custom Endpoints

The most common use case is configuring Ollama with custom base URLs.

Default Ollama Configuration

By default, GitWhisper expects Ollama at:

http://localhost:11434

Custom Port

If Ollama is running on a different port:

# Use custom port
gitwhisper commit --model ollama --base-url http://localhost:8080

# Set as default
gitwhisper set-defaults --model ollama --base-url http://localhost:8080

Remote Ollama Server

Access Ollama running on another machine:

# Use remote Ollama server
gitwhisper commit --model ollama --base-url http://192.168.1.100:11434

# Or use hostname
gitwhisper commit --model ollama --base-url http://ollama-server.local:11434

# Set as default
gitwhisper set-defaults --model ollama --base-url http://ollama-server.local:11434

Docker Ollama

If Ollama is in a Docker container with port mapping:

# Docker with custom port mapping
docker run -d -p 8080:11434 ollama/ollama

# Configure GitWhisper
gitwhisper set-defaults --model ollama --base-url http://localhost:8080

Setting Custom Endpoints

Command Line Usage

Use --base-url flag for one-time usage:

# Commit with custom endpoint
gitwhisper commit --model ollama --base-url http://custom-host:11434

# Analyze with custom endpoint
gitwhisper analyze --model ollama --base-url http://custom-host:11434

Set as Default

Save custom endpoint as default:

# Set default base URL for Ollama
gitwhisper set-defaults --model ollama --base-url http://localhost:8080

# Include model variant
gitwhisper set-defaults --model ollama --model-variant codellama --base-url http://localhost:8080

# View current configuration
gitwhisper show-defaults

Configuration File

Custom endpoints are stored in ~/.git_whisper.yaml:

defaults:
  model: ollama
  model_variant: codellama
  base_url: http://localhost:8080

Advanced Configurations

HTTPS Endpoints

Use HTTPS for secure connections:

# HTTPS endpoint
gitwhisper set-defaults --model ollama --base-url https://ollama.company.com:443

URL with Paths

If your API is behind a reverse proxy with a path:

# Endpoint with path
gitwhisper set-defaults --model ollama --base-url http://api.company.com/ai/ollama

Multiple Ollama Instances

Switch between different Ollama servers:

# Development server
gitwhisper commit --model ollama --base-url http://dev-ollama:11434

# Production server
gitwhisper commit --model ollama --base-url http://prod-ollama:11434

# Local instance
gitwhisper commit --model ollama --base-url http://localhost:11434

Network Configuration

Firewall Rules

Ensure your firewall allows connections:

Linux (iptables)
Linux (ufw)
macOS

# Allow incoming connections on Ollama port
sudo iptables -A INPUT -p tcp --dport 11434 -j ACCEPT

# Save rules
sudo iptables-save

# Allow Ollama port
sudo ufw allow 11434/tcp

# Check status
sudo ufw status

# macOS firewall settings
# Go to System Preferences > Security & Privacy > Firewall
# Add Ollama to allowed apps

Network Binding

Configure Ollama to accept remote connections:

# Start Ollama with specific bind address
OLLAMA_HOST=0.0.0.0:11434 ollama serve

# Or use environment variable
export OLLAMA_HOST=0.0.0.0:11434
ollama serve

Binding to 0.0.0.0 exposes Ollama to all network interfaces. Use firewall rules to restrict access in production.

Docker Compose Setup

Complete Docker Configuration

# docker-compose.yml
version: '3.8'

services:
  ollama:
    image: ollama/ollama
    ports:
      - "11434:11434"
    volumes:
      - ollama_data:/root/.ollama
    environment:
      - OLLAMA_HOST=0.0.0.0:11434
    restart: unless-stopped

volumes:
  ollama_data:

Start and configure:

# Start Ollama in Docker
docker-compose up -d

# Pull a model
docker-compose exec ollama ollama pull llama3.2

# Configure GitWhisper
gitwhisper set-defaults --model ollama --base-url http://localhost:11434

Kubernetes Deployment

Ollama in Kubernetes

# ollama-deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
  name: ollama
spec:
  replicas: 1
  selector:
    matchLabels:
      app: ollama
  template:
    metadata:
      labels:
        app: ollama
    spec:
      containers:
      - name: ollama
        image: ollama/ollama
        ports:
        - containerPort: 11434
        env:
        - name: OLLAMA_HOST
          value: "0.0.0.0:11434"
---
apiVersion: v1
kind: Service
metadata:
  name: ollama-service
spec:
  selector:
    app: ollama
  ports:
  - protocol: TCP
    port: 11434
    targetPort: 11434
  type: LoadBalancer

Deploy and use:

# Deploy to Kubernetes
kubectl apply -f ollama-deployment.yaml

# Get service endpoint
kubectl get svc ollama-service

# Configure GitWhisper with service endpoint
gitwhisper set-defaults --model ollama --base-url http://<EXTERNAL-IP>:11434

Proxy Configuration

Corporate Proxy Setup

If GitWhisper needs to go through a corporate proxy:

# Set HTTP proxy environment variables
export HTTP_PROXY=http://proxy.company.com:8080
export HTTPS_PROXY=http://proxy.company.com:8080

# Then use GitWhisper normally
gitwhisper commit --model ollama

Nginx Reverse Proxy

# /etc/nginx/sites-available/ollama
server {
    listen 80;
    server_name ollama.company.com;

    location / {
        proxy_pass http://localhost:11434;
        proxy_set_header Host $host;
        proxy_set_header X-Real-IP $remote_addr;
        proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
        proxy_set_header X-Forwarded-Proto $scheme;
    }
}

Configure GitWhisper:

gitwhisper set-defaults --model ollama --base-url http://ollama.company.com

Troubleshooting

Connection Refused

Error: Connection refused

Checklist:

Verify Ollama is running: curl http://localhost:11434/api/version
Check the port is correct
Verify firewall settings
Check Ollama logs: docker logs ollama (if using Docker)

Connection Timeout

Error: Connection timeout

Solutions:

Check network connectivity: ping <hostname>
Verify firewall allows connections
Check if Ollama is bound to correct interface
Try increasing timeout (if supported)

Invalid URL Format

Error: Invalid base URL

Format requirements:

Must include protocol: http:// or https://
Include port if not default: :11434
No trailing slash

Correct:

http://localhost:11434
https://ollama.company.com

Incorrect:

localhost:11434 (missing protocol)
http://localhost:11434/ (trailing slash)

Certificate Errors (HTTPS)

Error: SSL certificate verification failed

For self-signed certificates:

Add certificate to system trust store
Or use HTTP instead (if on trusted network)
Configure certificate validation (if GitWhisper supports it)

DNS Resolution Failed

Error: Could not resolve hostname

Solutions:

Check DNS: nslookup <hostname>
Use IP address instead: http://192.168.1.100:11434
Add to /etc/hosts if local server
Verify VPN connection if required

Security Considerations

Security Best Practices:

Use HTTPS for remote endpoints when possible
Restrict access with firewall rules
Use authentication if available
Avoid public exposure of Ollama servers
Use VPN for remote access to internal servers

Secure Remote Access

# Option 1: SSH Tunnel
ssh -L 11434:localhost:11434 user@remote-host
gitwhisper commit --model ollama --base-url http://localhost:11434

# Option 2: VPN
# Connect to VPN first, then:
gitwhisper commit --model ollama --base-url http://internal-ollama:11434

# Option 3: Tailscale/WireGuard
# Use private network addresses
gitwhisper commit --model ollama --base-url http://100.64.0.1:11434

Testing Custom Endpoints

Verify Connectivity

# Test if endpoint is reachable
curl http://localhost:11434/api/version

# Test with remote endpoint
curl http://remote-host:11434/api/version

# Test HTTPS endpoint
curl https://ollama.company.com/api/version

Test with GitWhisper

# Test custom endpoint
gitwhisper list-models --base-url http://localhost:8080

# If it works, set as default
gitwhisper set-defaults --model ollama --base-url http://localhost:8080

Common Scenarios

Scenario 1: Development Team

Setup: Shared Ollama server for team

# Server admin sets up Ollama
OLLAMA_HOST=0.0.0.0:11434 ollama serve

# Team members configure GitWhisper
gitwhisper set-defaults --model ollama --base-url http://dev-ai-server:11434

Scenario 2: CI/CD Pipeline

Setup: Ollama in CI environment

# .gitlab-ci.yml or .github/workflows/main.yml
services:
  - name: ollama/ollama
    alias: ollama

variables:
  OLLAMA_BASE_URL: "http://ollama:11434"

script:
  - gitwhisper commit --model ollama --base-url $OLLAMA_BASE_URL

Scenario 3: Home Lab

Setup: Ollama on home server

# Access from multiple machines
gitwhisper set-defaults --model ollama --base-url http://homeserver.local:11434

Configuration Examples

Example 1: Multi-Environment

# ~/.bashrc or ~/.zshrc
alias gw-local="gitwhisper commit --model ollama --base-url http://localhost:11434"
alias gw-dev="gitwhisper commit --model ollama --base-url http://dev-ollama:11434"
alias gw-prod="gitwhisper commit --model ollama --base-url http://prod-ollama:11434"

Example 2: Load Balanced

# Behind load balancer
gitwhisper set-defaults --model ollama --base-url http://ollama-lb.company.com

Next Steps

Ollama Guide

Learn more about using Ollama

Configuration

Configure default settings

Docker Setup

Official Ollama Docker image

Troubleshooting

Common issues and solutions

Supported Providers

Model Configuration

​Overview

Self-Hosted Ollama

Corporate Proxies

API Gateways

Development Testing

​Ollama Custom Endpoints

​Default Ollama Configuration

​Custom Port

​Remote Ollama Server

​Docker Ollama

​Setting Custom Endpoints

​Command Line Usage

​Set as Default

​Configuration File

​Advanced Configurations

​HTTPS Endpoints

​URL with Paths

​Multiple Ollama Instances

​Network Configuration

​Firewall Rules

​Network Binding

​Docker Compose Setup

​Complete Docker Configuration

​Kubernetes Deployment

​Ollama in Kubernetes

​Proxy Configuration

​Corporate Proxy Setup

​Nginx Reverse Proxy

​Troubleshooting

​Security Considerations

​Secure Remote Access

​Testing Custom Endpoints

​Verify Connectivity

​Test with GitWhisper

​Common Scenarios

​Scenario 1: Development Team

​Scenario 2: CI/CD Pipeline

​Scenario 3: Home Lab

​Configuration Examples

​Example 1: Multi-Environment

​Example 2: Load Balanced

​Next Steps

Ollama Guide

Configuration

Docker Setup

Troubleshooting

Build docs developers (and LLMs) love

Overview

Ollama Custom Endpoints

Default Ollama Configuration

Custom Port

Remote Ollama Server

Docker Ollama

Setting Custom Endpoints

Command Line Usage

Set as Default

Configuration File

Advanced Configurations

HTTPS Endpoints

URL with Paths

Multiple Ollama Instances

Network Configuration

Firewall Rules

Network Binding

Docker Compose Setup

Complete Docker Configuration

Kubernetes Deployment

Ollama in Kubernetes

Proxy Configuration

Corporate Proxy Setup

Nginx Reverse Proxy

Troubleshooting

Security Considerations

Secure Remote Access

Testing Custom Endpoints

Verify Connectivity

Test with GitWhisper

Common Scenarios

Scenario 1: Development Team

Scenario 2: CI/CD Pipeline

Scenario 3: Home Lab

Configuration Examples

Example 1: Multi-Environment

Example 2: Load Balanced

Next Steps