tomevault-io mo-arch - Code Architecture Query
Query code architecture using README_AI.md index files generated by codeindex.
Workflow
Step 1: Check Project Index
cat PROJECT_INDEX.md 2>/dev/null
If exists, this gives you the project overview with all modules listed.
Step 2: Find Directory Indexes
find . -name "README_AI.md" -type f | head -10
Step 3: Read Relevant Module Index
Based on user's question, locate the relevant module and read its README_AI.md:
# Example: user asks about parser
cat src/parser/README_AI.md
Step 4: Trace Dependencies
Check the Consumes and Provides sections in README_AI.md to understand module relationships.
Step 5: Deep Dive (if needed)
Only read actual source files when README_AI.md doesn't have enough detail.
Response Format
When answering architecture questions:
- Start with the module location - "The parser is in
src/codeindex/parser.py" - Explain its purpose - From the Purpose section
- Show relationships - From Consumes/Provides
- Provide code reference - Line numbers if relevant
Example
User: "Where is the parser implemented?"
Response:
Based on README_AI.md, the parser is in `src/codeindex/parser.py`.
**Purpose**: Tree-sitter based Python parser that extracts Symbol
(classes, functions, methods) and Import statements into ParseResult.
**Key exports**:
- `parse_file()` - Parse a single Python file
- `ParseResult` - Container for parsed data
- `Symbol` - Represents a code symbol
**Dependencies**: tree-sitter, tree-sitter-python
When No Index Exists
If README_AI.md files don't exist, suggest:
This directory is not indexed. Run:
codeindex scan ./path/to/dir
Or for the entire project:
codeindex list-dirs | xargs -P 4 -I {} codeindex scan {}
codeindex index
Converted and distributed by TomeVault — claim your Tome and manage your conversions.