1. Overview
The buildScripts/ai/queryKnowledgeBase.mjs script is a critical tool for the AI agent's ability to understand the Neo.mjs framework. This ticket covers enhancing the script to be more robust and powerful by refactoring its command-line interface and adding a content-type filtering feature.
2. Scope of Work
2.1. Refactor to Use commander
- Task: Replace the
yargs dependency and implementation in buildScripts/ai/queryKnowledgeBase.mjs with commander.
- Reason: This aligns the script with the conventions used in other project build tools (e.g.,
buildThemes.mjs), creating a more consistent development experience.
- Implementation:
- Import
Command from commander/esm.mjs.
- Set up the program with a name, version, and options.
2.2. Implement Content Type Filtering
- Task: Add a new command-line option to filter query results by content type.
- Implementation:
- Add a
--type (alias -t) option to the commander setup.
- The option should accept the following values:
all, blog, guide, src, example.
- The default value should be
all.
- After the initial database query, filter the results based on the
source file path in the metadata before the scoring algorithm is applied.
2.3. Update AGENTS.md Documentation
- Task: Update the AI agent guidelines to reflect the new functionality.
- Implementation:
- In the "The Query Command" section, update the example to show the new
--type flag.
- In the "Query Strategies" section, add examples and guidance on how to use the new flag for more targeted searches.
- Specifically mention the strategy of using
-t src or -t example when conceptual guides are insufficient or to find concrete implementation details.
3. Acceptance Criteria
- Running
npm run ai:query with the new flags works as expected.
- The script correctly filters results when a
--type other than all is specified.
AGENTS.md is updated with clear and accurate instructions on using the new feature.- The
yargs dependency is removed from the script.
1. Overview
The
buildScripts/ai/queryKnowledgeBase.mjsscript is a critical tool for the AI agent's ability to understand the Neo.mjs framework. This ticket covers enhancing the script to be more robust and powerful by refactoring its command-line interface and adding a content-type filtering feature.2. Scope of Work
2.1. Refactor to Use
commanderyargsdependency and implementation inbuildScripts/ai/queryKnowledgeBase.mjswithcommander.buildThemes.mjs), creating a more consistent development experience.Commandfromcommander/esm.mjs.2.2. Implement Content Type Filtering
--type(alias-t) option to thecommandersetup.all,blog,guide,src,example.all.sourcefile path in the metadata before the scoring algorithm is applied.2.3. Update
AGENTS.mdDocumentation--typeflag.-t srcor-t examplewhen conceptual guides are insufficient or to find concrete implementation details.3. Acceptance Criteria
npm run ai:querywith the new flags works as expected.--typeother thanallis specified.AGENTS.mdis updated with clear and accurate instructions on using the new feature.yargsdependency is removed from the script.