You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
samples/.prompts/code_freshness.md

63 lines
3.1 KiB

# Code Health and Style Guide Analysis
This document provides instructions for an AI assistant to analyze this repository against the Dart and Flutter style guide (`.prompts/llm.md`) and log opportunities for improvement.
## Workflow
The AI assistant must follow these steps exactly:
### 1. Setup
1. **Create Log Directory**: If it does not already exist, create a directory named `logs` in the repository root.
2. **Create Log File**: Create a new file inside the `logs` directory named `YYYY-MM-DD_HH-MM-SS-freshness-scores.md`, where `YYYY-MM-DD_HH-MM-SS` is the current timestamp.
### 2. Project Identification
1. **Find Projects**: Identify all sample projects by searching for `pubspec.yaml` files within the repository. Each directory containing a `pubspec.yaml` should be considered a separate project.
2. **Create a Queue**: Compile a list of the absolute paths to these project directories to process them one by one.
### 3. Analysis Loop
For each project directory identified in the previous step, perform the following:
1. **Check for Dart Files**:
* Verify that the project directory contains at least one `.dart` file.
* If not, add an entry to the log file stating that the project was skipped for this reason and proceed to the next project.
2. **Analyze Git History**:
* Run `git log` to find the most recent commit to a `.dart` file within that project's directory made by a human (i.e., not a bot).
* **Command**:
```bash
git log -1 --author='^(?!.*bot).*$' --pretty="format:%ad" --date=short -- ./**/*.dart
```
*(Note: This command should be run from within the project's directory).*
* **Handle No Commits**: If the command returns no output, note "No human commits found" for the log. If there are no commits at all, note that as well.
3. **Read and Assess Code**:
* Read all `.dart` files within the project directory (recursively).
* Compare the code against the rules and patterns defined in `.prompts/llm.md`.
* The assessment must focus on:
* Code organization and structure.
* Adherence to naming conventions.
* Proper use of Flutter and Dart patterns (e.g., `const` constructors, collection literals).
* State management best practices.
* Clarity, readability, and documentation.
4. **Log Findings**:
* Append a new entry to the log file using the following Markdown template. Ensure all fields are filled out.
```markdown
---
**Sample**: `path/to/sample`
**Last Human Commit**: `YYYY-MM-DD` or "No human commits found"
**Assessment Date**: `YYYY-MM-DD`
**Summary of Style Guide Adherence**:
A brief, one-paragraph summary of how well the project adheres to the style guide. Mention its strengths and weaknesses in general terms.
**Opportunities for Improvement**:
- A concrete, actionable bullet point describing a specific area for improvement.
- Another actionable bullet point.
- A third bullet point, if applicable. Focus on providing clear, specific, and helpful recommendations.
---
```