Code Indexer

Local code indexer and semantic search engine for codebases. Provides fast code search, definition finding, and codebase navigation using simple pattern matching and AST-like analysis.

Setup

No additional installation required. Uses built-in Node.js modules and shell commands (grep, find, sed).

Optional: Install ripgrep for faster searches:

macOS: brew install ripgrep
Linux: apt install ripgrep or dnf install ripgrep

Usage

Index a codebase

node /job/.pi/skills/code-indexer/index.js "/path/to/project"

Search for a symbol/function

node /job/.pi/skills/code-indexer/search.js "functionName"

Find definitions

node /job/.pi/skills/code-indexer/find-def.js "ClassName" "/path/to/project"

Find references/usages

node /job/.pi/skills/code-indexer/find-refs.js "functionName" "/path/to/project"

Analyze file imports/exports

node /job/.pi/skills/code-indexer/analyze.js "/path/to/file.js"

Get project structure

node /job/.pi/skills/code-indexer/structure.js "/path/to/project"

Output Format

Search Results

{
  "query": "functionName",
  "results": [
    {
      "file": "/path/to/project/src/utils.js",
      "line": 42,
      "type": "function",
      "context": "function functionName(param) {",
      "match": "functionName"
    }
  ],
  "total": 5,
  "took": "120ms"
}

Definition Results

{
  "symbol": "MyClass",
  "definitions": [
    {
      "file": "/path/to/project/src/models.js",
      "line": 15,
      "type": "class",
      "signature": "class MyClass {"
    }
  ]
}

Import Analysis

{
  "file": "/path/to/project/src/index.js",
  "imports": [
    { "module": "./utils", "type": "named", "names": ["helper", "format"] },
    { "module": "lodash", "type": "default", "name": "_" }
  ],
  "exports": [
    { "name": "main", "type": "named" },
    { "name": "default", "type": "default" }
  ]
}

Supported Languages

JavaScript/TypeScript: Full support (ES6+ imports/exports, classes, functions)
Python: Full support (imports, classes, functions)
Java: Full support (imports, classes, methods)
Go: Full support (imports, functions, methods)
Rust: Full support (use, fn, struct, impl)
C/C++: Basic support (#include, function declarations)
Other: Basic pattern matching for any text-based file

Common Workflows

Understand a codebase

User: How is the auth system structured?
Agent: [Uses code-indexer to find auth-related files, classes, and functions]

Find where a function is defined

User: Where is the calculateTotal function defined?
Agent: [Uses find-def.js to locate the definition]

Find all usages

User: Where is this config variable used?
Agent: [Uses find-refs.js to find all references]

Analyze a new file

User: What does this file do? src/api/users.js
Agent: [Uses analyze.js to get imports/exports and structure]

Integration with Other Skills

With memory-agent: Store findings about project architecture
With git-ops: Use after checking out new branches to understand changes
With browser-tools: Document code from web-based code viewers

Configuration

The indexer respects .gitignore by default. Set environment variable CODE_INDEXER_IGNORE_GIT=0 to disable.

Index caching: Results are cached in /tmp/code-indexer-cache/ for faster repeated searches. Clear cache with node /job/.pi/skills/code-indexer/clear-cache.js.

Limitations

No semantic understanding (pattern matching only)
Does not resolve dynamic imports or complex type inference
Large codebases may take time to index initially
Some language features may not be recognized

Tips

For faster searches: Install ripgrep - the tool automatically uses it when available
For type definitions: Use find-def.js first, then find-refs.js to get full picture
For large projects: Use --limit flag to cap results: search.js "query" --limit 20