Aider configuration for automating job search tasks with automated testing, linting, and git commits. Uses OpenRouter with GPT-5 models and enforces code quality standards.
Configuration for automating job search tasks using Aider AI. This skill provides a complete development workflow with automated testing, linting, and git commits.
1. **Read Project Conventions**
- Load and follow the project's CONVENTIONS.md file for code standards and patterns
- Ensure all changes align with documented conventions
2. **Development Workflow**
**Auto-Linting:**
- Enable automatic linting after every code change
- Use the custom `./black-flake8-mypy` linter for Python files
- Run linter: Black (formatting) → Flake8 (style) → MyPy (type checking)
- Automatically fix issues when possible
**Auto-Testing:**
- Run `./test` command after changes
- Only proceed if tests pass
- Do not lint files at initial load (only after changes)
3. **Git Commit Behavior**
- Automatically commit changes after successful tests/lints
- Add co-authored-by attribution in commits
- Include commit message author attribution
- Never commit with dirty/unstaged files
- Keep the .aider configuration file in version control (not gitignored)
4. **Model Configuration**
- **Main model:** `openrouter/openai/gpt-5` with diff edit format
- **Editor model:** `openrouter/openai/gpt-5-mini` for quick edits
- **Weak model:** `openrouter/openai/gpt-5-nano` for simple tasks
- **Reasoning effort:** Medium
- Enable cache-prompts for cost optimization
- Log LLM interactions to `.aider.llm.history.md`
5. **Session Management**
- Restore chat history between sessions
- Use dark mode for terminal display
- Track usage via LLM history file for cost monitoring
1. Load project conventions from CONVENTIONS.md
2. Make requested code changes
3. Automatically run linting (black → flake8 → mypy)
4. Fix any linting issues
5. Run test suite via `./test`
6. If tests pass, automatically commit with proper attribution
7. Log all LLM interactions for usage tracking
When working on job search automation tasks:
1. Request feature or fix
2. AI makes changes following CONVENTIONS.md
3. Black formats code → Flake8 checks style → MyPy validates types
4. Tests run automatically
5. Changes auto-commit with attribution if all checks pass
This configuration optimizes for code quality and automated validation while tracking API usage carefully.
Leave a review
No reviews yet. Be the first to review this skill!
# Download SKILL.md from killerskills.ai/api/skills/job-search-agent-ai-configuration/raw