Shellm Aider Configuration

Configuration for using Aider with the shellm project - a one-file Ollama CLI client written in bash.

What This Does

This skill configures Aider to work optimally with the shellm bash project by:

Configuration Details

**Key Settings:**

**Test Integration:**

Instructions

1. **Setup**: Place this configuration as `.aider.conf.yml` in the shellm project root

2. **What Gets Loaded**:

- AGENTS.md documentation for context

- shellm main file for editing

- Custom test runner script

3. **Workflow**:

- Aider will use architect mode for structured edits

- After each change, tests run automatically

- Manual git commits give you control

- Test failures block changes until resolved

4. **Model Configuration**:

- Thinking tokens set to 0 (disabled for performance)

- Model can be customized via `--model` flag

- Supports weak/editor model configuration

5. **Testing**:

- Test command executes `./run_tests.sh`

- Auto-test ensures quality gates

- Use `--test` flag to run tests and fix issues

Usage Examples

**Start Aider with this config:**

```bash

aider

```

**Override model:**

```bash

aider --model gpt-4o

```

**Disable auto-test temporarily:**

```bash

aider --no-auto-test

```

**Run tests and fix issues:**

```bash

aider --test

```

Important Notes

Customization

You can override any setting by:

1. Editing `.aider.conf.yml` directly

2. Using command-line flags (higher precedence)

3. Setting environment variables

Common customizations: