Aider Configuration with Local API

This skill configures Aider to work with a local OpenAI-compatible API server (like GitHub Copilot's local endpoint) while maintaining project conventions through a dedicated file.

Configuration

The configuration enables:

Setup Instructions

1. **Create the Aider configuration file** (`.aider.conf.yml` in your project root):

```yaml

dark-mode: true

read: CONVENTIONS.md

Ollama config (commented out)

#model: ollama/mistral-nemo

Copilot config

openai-api-key: dummy_key

4: true # Use gpt-4-0613 model for the main chat

openai-api-base: http://localhost:51234/v1

Claude config (commented out)

sonnet: true

```

2. **Create a `CONVENTIONS.md` file** in your project root to define coding standards, patterns, and guidelines that Aider should follow.

3. **Ensure your local API server is running** on `http://localhost:51234/v1` before starting Aider.

Usage

Run Aider from your project directory:

```bash

aider

```

Aider will automatically:

Alternative Backends

The configuration includes commented-out alternatives:

**Ollama** (local open-source models):

```yaml

model: ollama/mistral-nemo

```

**Claude** (Anthropic API):

```yaml

sonnet: true

```

Uncomment and configure as needed for your preferred backend.

Notes