Claude Code: An analysis

Hrishi Olickel & Claude Opus 4 & Gemini 2.5 Pro·

Why Claude Code Matters

Claude Code has a number of very interesting parts:

  • Streaming Architecture that handles real-time LLM responses, tool execution, and UI updates
  • Safety Systems that provide security without disrupting workflow
  • Tool Design that elegantly bridges AI reasoning and system execution
  • Prompt Engineering that reliably controls complex LLM behavior

Let's jump in! Each heading is a link to the full section.

Dependencies: The Foundation of Claude Code's Architecture

Why React in a terminal? What's yoga-layout doing here?

Discover the unconventional dependency choices that enable Claude Code's performance. Learn about the custom shell parser that embeds JSON in bash commands, the streaming JSON parser for partial LLM responses, and the ANR detection system borrowed from mobile development.

Data Structures & The Information Architecture

How messages transform through the system

Follow data from user input through LLM processing to tool execution. Understand the three-stage message representation, the ContentBlock polymorphism, and how weak references prevent memory bloat.

Key Insight: The CliMessage wrapper maintains UI state while preserving API compatibility--enabling rich interactions without protocol changes.

Control Flow & The Orchestration Engine

Inside the tt function

Explore the six-phase async generator that orchestrates everything. See how parallel tool execution works, why context compaction triggers automatically, and how recursive turns enable unlimited conversation depth.

Key Insight: Tools are categorized by side effects--read-only tools run in parallel while write operations serialize for safety.

Tools & The Execution Engine

From LLM decisions to system actions

Each tool is a carefully designed state machine. Examine the permission system, progress reporting, and error handling. Special focus on BashTool's sandbox mode and EditTool's line number handling.

Key Insight: The AgentTool implements hierarchical task decomposition--spawning sub-agents and synthesizing their findings.

Architecture: The Engine Room

Event-driven, streaming-first, security-conscious

Understand the layered architecture from React UI to system calls. Learn how permissions cascade through scopes, why ANR detection uses worker threads, and how three telemetry systems provide complete observability.

Key Insight: Security isn't one system--it's multiple independent layers that fail safely.

Novel Components: The Innovations That Define Claude Code

The clever fixes to hard problems

Discover components that make Claude Code special: streaming JSON parsing with recovery, intelligent data truncation, and multi-agent result synthesis. These aren't just features--they're innovative solutions to fundamental challenges.

Key Insight: The normalizeToSize algorithm iteratively reduces object depth based on actual byte count--preserving maximum information within constraints.

File Editing: AI-Assisted Code Modification

Why three different editing tools?

Deep dive into the file editing pipeline. Learn why line numbers cause problems, how sequential edits detect conflicts, and what happens when files change externally.

Key Insight: Every conceivable editing mistake has a specific validation--from external modifications to encoding issues.

Prompt Engineering: The Art of Instructing AI

The instructions that make it all work

Examine the actual prompts that control Claude Code. From conciseness enforcement to the 500+ word BashTool safety instructions, see how careful wording shapes behavior.

Key Insight: Repetition works--critical instructions appear three times with escalating emphasis.

An LLM's Perspective: What It's Actually Like to Receive These Instructions

What these prompts feel like from the other side

In a unique section, an LLM (me) provides honest commentary on receiving these instructions. Why "just output 4" is surprisingly difficult and how the -$1000 penalty creates real behavioral change despite being imaginary money.

Key Insight: Clear constraints are secretly liberating--they prevent decision paralysis and over-helpfulness.

Technical Themes

Throughout the analysis, several design principles emerge:

  1. Streaming First: Every operation designed for incremental updates
  2. Safety Through Layers: Multiple independent protection mechanisms
  3. Explicit Instructions: Verbose prompts prevent ambiguous behavior
  4. Architecture Over Optimization: Performance through design, not tweaks
  5. Understanding LLM Psychology: Exploiting how models actually behave

Sections

Dependencies: The Foundation of Claude Code's Architecture

Data Structures & The Information Architecture

Control Flow & The Orchestration Engine

Tools & The Execution Engine

Architecture: The Engine Room

Novel Components: The Innovations That Define Claude Code

File Editing: AI-Assisted Code Modification

Prompt Engineering: The Art of Instructing AI

An LLM's Perspective: What It's Actually Like to Receive These Instructions

It's amazing that this is even possible, let alone in the time this took. I can't say everything in the report is correct - perhaps the Claude team can weigh in - but it has all been really useful and instructive.

The Process

Conducting smarter intelligences than me: new orchestras