claude-code-source-analysis (xiaonancs)

A systems-engineering reconstruction of the complete Claude Code agent harness from a reverse-engineered v2.1.88 snapshot (includes the extracted source zip plus appendix analyses). Part of a series that also covers OpenAI Codex CLI internals, giving a comparative harness view.

View source on GitHub

Key takeaways

• 01

  Reconstructs the full harness from an actual extracted v2.1.88 snapshot

• 02

  Sister studies (Codex CLI) enable cross-harness comparison

• 03

  Strong on context management and compaction mechanics

Flows built on this research

Memory & Context

Implement Context Compaction

Keep long agent sessions alive by compacting old conversation into structured summaries before the context window fills.

4 steps · 90-120 minutes

Memory & Context

Session Summaries That Survive Compaction

Write durable session summaries at the right moments so knowledge outlives compaction and process restarts.

4 steps · 60-90 minutes

Memory & Context

Prompt-Cache-Aware Conversation Design

Structure your agent's context for prefix caching: stable prefixes, append-only history, and cache-friendly dynamic content.

3 steps · 45-75 minutes

MCP & Tooling

Tool Result Size Management

Stop tool outputs from flooding context: budgets per tool, smart truncation, pagination, and reference-based results.

4 steps · 60-90 minutes