Problem:
The main problem is that Qoder’s Repo Wiki is stored externally, making it inaccessible to AI agents and difficult to integrate with standard development workflows. This prevents AI from using the wiki as a valuable source of context to improve its functionality.
Solution
How should it work?
For a more integrated and powerful experience, Qoder’s Repo Wiki should work as follows:
- In-Workspace Generation: The core wiki content, generated automatically by Qoder’s AI, should be created as Markdown files within a designated folder in the user’s project directory (e.g., /docs or /wiki).
- Version Control: This wiki folder should be part of the project’s Git repository. This allows the wiki to be versioned alongside the code, ensuring documentation always matches the codebase it describes. It also makes it easy to sync the wiki with services like GitHub Wiki.
- AI Context Access: The AI agents (Chat, Agent, and CLI) in Qoder should have direct read access to these in-workspace markdown files. They would use this information as a primary source of context, alongside the code itself, when performing tasks.
- Bidirectional Sync: The system should not just be a one-time generation. Any manual edits made by developers to the wiki files should be read by the AI to update its understanding of the project. Conversely, the AI could be prompted to update the wiki with new information based on recent code changes or feature implementations.
- GitHub Integration: Qoder should provide a streamlined way to link the in-workspace wiki folder to a GitHub Wiki, allowing for easy push/pull synchronization. This would enable collaboration and make the documentation accessible outside of the Qoder IDE.
In short, the wiki should move from an external, isolated database to a first-class, version-controlled part of the project itself, directly accessible to the AI.
Use Case
Use Cases
- Policy-Compliant Code Generation: An AI can generate new code that adheres to internal standards defined in the wiki. For example, if a security policy in the wiki requires all API calls to be logged, the AI will automatically include the necessary logging functions when a developer asks it to create an API endpoint. This prevents security vulnerabilities and ensures consistency across the codebase.
- Contextual Refactoring: When a developer asks the AI to refactor a part of the codebase, the AI can read architectural decisions and design patterns documented in the wiki. This allows it to make more informed and intelligent refactoring choices that maintain the project’s intended structure and performance characteristics, rather than simply optimizing for code readability.
- Intelligent Onboarding and Support: New team members can ask the AI complex questions about the project’s architecture, design philosophy, and module dependencies. The AI can provide synthesized answers by referencing the wiki’s documentation, along with the codebase itself, significantly speeding up the onboarding process and providing immediate support for project-specific queries.
When would you use this?
Priority
-
High - Blocking issue -
Medium - Important improvement -
Low - Nice to have
Additional Info
Awesome Github Wikis