AI Catalyst 24/7 Automation Configuration
Production-ready Aider configuration optimized for continuous automated operation. This configuration prioritizes reliability, cost-effectiveness, and clean automation workflows over interactive features.
Overview
This skill configures Aider for 24/7 automated operation with:
Automatic commit handlingCost-effective model selection (DeepSeek)Performance optimizations for long-running processesSafety settings for unattended operationReduced verbosity for cleaner logsConfiguration Settings
Automation Mode
**auto-commits**: Enabled - automatically commit changes without prompting**auto-test**: Disabled - testing handled separately in validation pipeline**dirty-commits**: Enabled - allow commits even with uncommitted changes**commit-prompt**: Disabled - no interactive commit message prompts**stream**: Disabled - cleaner logs for automation**show-diffs**: Disabled - reduce log noise in automation modePerformance Optimization
**map-tokens**: 1024 - balanced repository map size**cache-prompts**: Enabled - improve response time and reduce costs**max-chat-history-tokens**: 4096 - limit context for cost control**restore-chat-history**: Disabled - fresh context for each automation runModel Selection
**Primary model**: `openrouter/deepseek/deepseek-chat` - cost-effective for main tasks**Weak model**: `openrouter/openai/gpt-3.5-turbo` - for simple operations**Editor model**: `openrouter/deepseek/deepseek-chat` - consistent cost-effective editingSafety & Reliability
**lint**: Disabled - handled in separate validation step**test-cmd**: Empty - testing managed by workflow**retry-delay**: 60 seconds - backup rate limiting**max-retries**: 3 - prevent infinite retry loopsGit Integration
**git-diffs**: Enabled - track changes**gitignore**: Enabled - respect repository ignore rulesOutput Settings
**pretty**: Disabled - consistent formatting for automation**verbose**: Disabled - reduce log verbosity**no-pretty**: Enabled - disable ANSI colors and formatting**dark-mode**: Disabled - consistent output rendering**analytics**: Disabled - no telemetry in automationUsage
Create a `.aider.conf.yml` file in your repository root with these settings:
```yaml
Automation Settings
auto-commits: true
auto-test: false
check-update: false
dirty-commits: true
pretty: false
stream: false
show-diffs: false
Performance Settings
map-tokens: 1024
cache-prompts: true
max-chat-history-tokens: 4096
Safety Settings
lint: false
test-cmd: ""
commit-prompt: false
Model Settings
model: openrouter/deepseek/deepseek-chat
weak-model: openrouter/openai/gpt-3.5-turbo
Git Settings
git-diffs: true
gitignore: true
Output Settings
verbose: false
no-pretty: true
dark-mode: false
Editor Settings
editor-model: openrouter/deepseek/deepseek-chat
editor-edit-format: diff
Advanced Settings
restore-chat-history: false
analytics: false
check-update: false
upgrade: false
Rate Limiting
retry-delay: 60
max-retries: 3
```
Integration with CI/CD
This configuration is designed to work with automated workflows:
1. **Pre-commit hooks**: Disable - auto-commits handle changes
2. **Linting**: Run as separate workflow step after Aider
3. **Testing**: Execute in dedicated validation phase
4. **Rate limiting**: Handled at workflow level; config provides backup
Cost Optimization
Uses DeepSeek models by default (significantly cheaper than GPT-4)Limits chat history tokens to prevent context bloatEnables prompt caching for repeated operationsDisables unnecessary features (analytics, update checks)Best Practices
Run linting and testing as separate workflow stepsMonitor token usage via workflow metricsAdjust `map-tokens` based on repository sizeUse workflow-level rate limiting as primary controlReview auto-commits periodically for qualityConstraints
Not suitable for interactive development sessionsRequires external linting/testing pipelineAssumes OpenRouter API access for model routingMinimal error output - check workflow logs for debuggingRelated Resources
[Aider documentation](https://aider.chat/docs/)[OpenRouter model pricing](https://openrouter.ai/models)AI Catalyst repository: AccidentalJedi/AI-Catalyst