Qwen3-Coder Aider Configuration

Comprehensive Aider configuration optimized for Qwen3-Coder, the code-specialized version of Qwen3 LLM. This configuration covers all available Aider settings including model selection, git integration, output customization, and automated workflows.

What This Skill Does

Provides a complete `.aider.conf.yml` template with all configuration options for Aider, the AI pair programming tool. Includes settings for:

Instructions

1. Locate Configuration Directory

Determine where to place your configuration file:

2. Create Configuration File

Create `.aider.conf.yml` with your selected options. Uncomment and set values for the features you need:

**Basic Model Setup:**

```yaml

Specify custom model

model: qwen/qwen3-coder

API configuration

openai-api-base: https://your-api-endpoint.com

openai-api-key: your-key-here

Edit format (options: whole, diff, udiff)

edit-format: whole

```

**Git Integration:**

```yaml

Enable git features

git: true

gitignore: true

auto-commits: true

dirty-commits: true

Commit attribution

attribute-author: true

attribute-committer: true

attribute-commit-message-author: true

```

**Output Customization:**

```yaml

Dark mode colors

dark-mode: true

pretty: true

stream: true

Custom colors (hex values)

user-input-color: "#00cc00"

assistant-output-color: "#0088ff"

tool-error-color: "#FF2222"

Code theme (default, monokai, solarized-dark, solarized-light)

code-theme: monokai

```

**Linting and Testing:**

```yaml

Automatic linting

auto-lint: true

lint-cmd:

- "python: ruff check --fix"

- "javascript: eslint --fix"

Automatic testing

auto-test: true

test-cmd: pytest

```

**Performance Optimization:**

```yaml

Repo map configuration

map-tokens: 2048

map-refresh: auto

map-multiplier-no-files: 2

Prompt caching

cache-prompts: true

cache-keepalive-pings: 3

Chat history limits

max-chat-history-tokens: 4096

```

3. Set API Keys Securely

For sensitive keys, use `.env` file instead of YAML:

```bash

Create .env in git root

echo "OPENAI_API_KEY=your-key" > .env

echo "ANTHROPIC_API_KEY=your-key" >> .env

Specify in config

env-file: .env

```

4. Configure Model Settings

For custom or unknown models, create model settings files:

**Model Settings (`.aider.model.settings.yml`):**

```yaml

edit_format: whole

weak_model_name: qwen/qwen3-coder-mini

use_repo_map: true

send_undo_reply: false

```

**Model Metadata (`.aider.model.metadata.json`):**

```json

{

"qwen/qwen3-coder": {

"max_tokens": 8192,

"max_input_tokens": 32768,

"input_cost_per_token": 0.0000001,

"output_cost_per_token": 0.0000002

}

}

```

Reference in config:

```yaml

model-settings-file: .aider.model.settings.yml

model-metadata-file: .aider.model.metadata.json

```

5. Configure Language-Specific Linting

Set up linting for multiple languages:

```yaml

lint-cmd:

- "python: ruff check --fix --select=E,F,W"

- "javascript: eslint --fix --max-warnings=0"

- "typescript: eslint --fix --ext .ts,.tsx"

- "go: gofmt -w"

- "rust: cargo fmt"

```

6. Set Up Read-Only Context Files

Provide reference files without allowing edits:

```yaml

read:

- README.md

- docs/architecture.md

- .env.example

```

7. Configure History Management

```yaml

Input history (command recall)

input-history-file: .aider.input.history

Chat history (conversation log)

chat-history-file: .aider.chat.history.md

restore-chat-history: true

LLM conversation log (debugging)

llm-history-file: .aider.llm.history

```

8. Optimize for Large Repositories

```yaml

Only consider current subtree

subtree-only: true

Use .aiderignore to exclude directories

aiderignore: .aiderignore

Disable model warnings for speed

show-model-warnings: false

```

Create `.aiderignore`:

```

node_modules/

dist/

build/

*.min.js

vendor/

```

9. Verify Configuration

Test your configuration:

```bash

Show active configuration

aider --verbose

Show system prompts (debug)

aider --show-prompts

Show repo map (debug)

aider --show-repo-map

Dry run (no file changes)

aider --dry-run --message "Explain this codebase"

```

10. Use Configuration in Commands

Override config values at runtime:

```bash

Use different model temporarily

aider --model gpt-4o

Disable auto-commits for this session

aider --no-auto-commits

Enable verbose output

aider --verbose

Apply changes from file

aider --apply changes.md

```

Usage Examples

**Example 1: Project-Specific Python Config**

```yaml

.aider.conf.yml

model: qwen/qwen3-coder

dark-mode: true

auto-lint: true

lint-cmd:

- "python: ruff check --fix"

auto-test: true

test-cmd: pytest tests/

auto-commits: true

commit-prompt: "Brief summary of changes for Python project"

```

**Example 2: Multi-Language Monorepo**

```yaml

model: gpt-4o

subtree-only: true

lint-cmd:

- "python: ruff check --fix"

- "javascript: eslint --fix"

- "typescript: eslint --fix --ext .ts"

map-tokens: 4096

cache-prompts: true

```

**Example 3: High-Performance Setup**

```yaml

model: qwen/qwen3-coder

cache-prompts: true

cache-keepalive-pings: 5

map-tokens: 2048

stream: true

show-model-warnings: false

```

Important Notes

Configuration Hierarchy

Aider searches for config files in this order:

1. `--config` flag path

2. `.aider.conf.yml` in current directory

3. `.aider.conf.yml` in git root

4. `~/.aider.conf.yml` in home directory

Command-line flags override all config file settings.