grepai-trace-callees_skill

Overview

This skill finds all functions called by a specific function using GrepAI trace callees. It helps quickly surface direct callees, their file locations, and code context so you can understand behavior and dependencies at a glance. Results can be returned in human-readable or JSON formats for scripting and documentation.

How this skill works

The skill runs grepai trace callees against a provided function name and inspects static call sites to list direct callees, method calls, and package-level calls. It supports fast and precise extraction modes; precise mode improves accuracy using tree-sitter when available. Output includes file paths, line numbers, callee names, and surrounding context, with optional compact JSON for AI workflows.

When to use it

• Map what a function does during code review or refactoring
• Discover direct dependencies and side effects of a function
• Generate dependency lists for documentation or change logs
• Script checks to see if a function calls specific APIs or DB methods
• Quickly identify functions with many callees as potential code smells

Best practices

• Start with --mode fast for quick results, switch to --mode precise for accuracy
• Combine trace callees with trace callers or trace graph for full impact analysis
• Filter JSON output (jq) to focus on file types or exclude tests
• Use compact JSON for downstream AI consumption or automation
• Ensure your code index and symbols are up to date before running

Example use cases

• List callees of ProcessOrder to document its behavior and side effects
• Check whether processOrder calls sendEmail using a jq filter in CI
• Count database-related callees to assess data-layer coupling
• Generate a reusable markdown dependency section for onboarding docs
• Run recursive analysis with trace graph for deeply nested dependencies