TLDR Code Analysis

Analyze code efficiently with 95% token savings.

Available Commands

5-Layer Analysis

Layer 1: AST         → ~500 tokens  (signatures, imports, classes)
Layer 2: Call Graph  → +440 tokens  (cross-file relationships)
Layer 3: CFG         → +110 tokens  (control flow, complexity)
Layer 4: DFG         → +130 tokens  (data flow, variables)
Layer 5: PDG         → +150 tokens  (program slicing)
────────────────────────────────────────────────────────
Total: ~1,200 tokens vs 23,000 raw = 95% savings

Quick Start

# First, check if tldr is installed
which tldr || echo "Install with: pip install tldr-code"

# Explore a codebase
tldr tree src/
tldr structure src/ --lang python

# Get context for a function
tldr context main --project src/ --depth 2

# Before refactoring - check impact
tldr impact process_data src/

Installation

The tldr CLI must be installed separately:

pip install tldr-code
# Or via uv:
uv pip install tldr-code

When to Use

• Before reading files: Get structure first, read selectively
• Understanding functions: Use context for LLM-ready summaries
• Debugging: Use slice to find what affects a buggy line
• Refactoring: Use impact before changing functions
• Code review: Use arch to detect layer violations