Aider Configuration Template

A complete `.aider.conf.yml` configuration template from Chapman University's CPSC 298 (LLM Coding Techniques) course. This file documents all valid Aider configuration entries for setting up AI-powered coding workflows.

What This Skill Provides

This configuration template enables you to:

Instructions

When a user requests Aider configuration help, follow these steps:

1. **Assess Requirements**

- Ask what LLM provider they're using (OpenAI, Anthropic, other)

- Determine if they need git auto-commits enabled

- Check if they want linting/testing automation

- Identify any custom requirements (dark mode, caching, etc.)

2. **Generate Configuration**

- Create a `.aider.conf.yml` file at the git repository root or home directory

- Include only the relevant configuration options (don't copy all commented defaults)

- Use the template structure: category headers with specific options below

3. **Essential Configuration Options**

**API Keys** (choose one or more):

```yaml

openai-api-key: your-key-here

anthropic-api-key: your-key-here

```

**Model Selection**:

```yaml

model: claude-3-5-sonnet-20240620 # or gpt-4o, etc.

```

**Git Integration**:

```yaml

auto-commits: true # Auto-commit LLM changes

dirty-commits: true # Allow commits when repo is dirty

attribute-author: true # Credit aider in author field

```

**Linting**:

```yaml

auto-lint: true

lint-cmd: ["python: ruff check --fix", "javascript: eslint --fix"]

```

**Testing**:

```yaml

auto-test: false

test-cmd: pytest

```

**Caching** (for cost savings):

```yaml

cache-prompts: true

cache-keepalive-pings: 6 # Keep cache warm for 30 minutes

```

4. **Common Use Cases**

**For Students/Learners**:

- Enable dark-mode/light-mode based on preference

- Set `show-diffs: true` to see changes

- Use `weak-model: gpt-4o-mini` for cheaper summarization

**For Production Work**:

- Set up `lint-cmd` for your languages

- Configure `test-cmd` for your test suite

- Enable `cache-prompts: true` for cost optimization

- Use `dry-run: true` when testing changes

**For Team Projects**:

- Set `attribute-commit-message-author: true` for transparency

- Configure `aiderignore` to exclude sensitive files

- Use `subtree-only: true` if working on monorepo subdirectory

5. **Configuration Precedence**

- Command-line flags override config file

- `.aider.conf.yml` in git root overrides home directory

- Use `--config` flag to specify alternate location

6. **API Key Security**

- Recommend using `.env` file for API keys instead of YAML

- Create `.env` in git root: `OPENAI_API_KEY=sk-...`

- Add `.env` to `.gitignore`

- Reference: https://aider.chat/docs/config/dotenv.html

7. **Validation**

- After creating config, test with: `aider --help` (should load without errors)

- Use `aider --show-prompts` to verify model settings

- Use `aider --show-repo-map` to test file detection

Key Configuration Categories

Examples

**Minimal OpenAI Setup**:

```yaml

model: gpt-4o

auto-commits: true

dark-mode: true

```

**Anthropic with Caching**:

```yaml

model: claude-3-5-sonnet-20240620

cache-prompts: true

cache-keepalive-pings: 6

auto-lint: true

lint-cmd: "python: ruff check --fix"

```

**Python Project with Testing**:

```yaml

model: gpt-4o-mini

auto-commits: true

auto-lint: true

lint-cmd: "python: ruff check --fix"

auto-test: false

test-cmd: pytest -v

show-diffs: true

```

Notes

Constraints