Recursive Decomposition Skill

Handle long-context tasks with Claude Code through recursive decomposition

What It Does • Installation • Usage • How It Works • Benchmarks • Acknowledgments

---

The Problem

When analyzing large codebases, processing many documents, or aggregating information across dozens of files, Claude's context window becomes a bottleneck. As context grows, **"context rot"** degrades performance:

• Missed details in long documents
• Decreased accuracy on information retrieval
• Hallucinated connections between distant content
• Degraded reasoning over large evidence sets

The Solution

This skill implements **Recursive Language Model (RLM)** strategies from [Zhang, Kraska, and Khattab's 2025 research](https://arxiv.org/abs/2512.24601), enabling Claude Code to handle inputs **up to 2 orders of magnitude beyond normal context limits**.

Instead of cramming everything into context, Claude learns to:

1. Filter — Narrow search space before deep analysis
2. Chunk — Partition inputs strategically
3. Recurse — Spawn sub-agents for independent segments
4. Verify — Re-check answers on smaller, focused windows
5. Synthesize — Aggregate results programmatically

---

What It Does

Task Type	Without Skill	With Skill
Analyze 100+ files	Context overflow / degraded results	Systematic coverage via decomposition
Multi-document QA	Missed information	Comprehensive extraction
Codebase-wide search	Manual iteration	Parallel sub-agent analysis

...