claude-devtools-inspector

Inspect Claude Code JSONL session logs to recover tool calls, token usage, and context compaction.

4|Updated May 16, 2026

Search Tags:#claude #token analysis #context window #tool calls #agent observability #jsonl #session debugging

Ready to run

npx skills add https://github.com/Aradotso/devtools-skills --skill claude-devtools-inspector

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: claude-devtools-inspector
Source: https://github.com/Aradotso/devtools-skills/tree/main/skills/claude-devtools-inspector
Command: npx skills add https://github.com/Aradotso/devtools-skills --skill claude-devtools-inspector

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Claude Code hides the details of what happened during a run, making it hard to debug tool behavior, understand token usage, or explain why context was compacted and actions changed.

Core Features & Use Cases

Full session inspection: Reconstructs Claude Code sessions from local ~/.claude/ JSONL logs into readable transcripts.
Tool call visibility: Displays exact tool inputs/outputs for actions like Read, Edit, Bash, Search, and more.
Token attribution & context analysis: Breaks down tokens by categories and visualizes context window compaction to show what was dropped.
Subagent and memory visibility: Shows nested agent trees with cost/duration and browses project memory layers.
Export and troubleshooting: Exports sessions to Markdown/JSON and includes guidance for common issues like missing logs or Docker mounts.

Quick Start

Open claude-devtools after launching Claude Code with a named session so the session logs appear in the claude-devtools sidebar from ~/.claude/sessions/.

Frequently Asked Questions about claude-devtools-inspector

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I inspect Claude Code session logs to see what happened during a run?▼

Inspect Claude Code session logs by reading JSONL files from ~/.claude/sessions/ to reconstruct readable transcripts. This recovers hidden execution details like exact tool inputs, thinking data, and nested subagent activity without modifying the original logs.

Can I analyze token usage and context window compaction from JSONL logs?▼

Analyze token usage and context window compaction by reconstructing per-turn data from JSONL logs. The tool breaks down tokens by category and visualizes context windows to show exactly what content was dropped during compaction.

How do I debug tool calls and agent behavior in prior Claude Code runs?▼

Debug tool calls and agent behavior by recovering exact tool inputs and outputs from prior runs. Reading local session logs allows you to audit actions like Read, Edit, Bash, and Search to understand why specific execution paths were taken.

What is the best way to export Claude session transcripts to Markdown or JSON?▼

The best way to export Claude session transcripts to Markdown or JSON is by reconstructing the session data from local JSONL logs. This process extracts tool calls, token attributions, and nested agent trees into structured formats for external review.

Why are my Claude Code session logs missing or not showing up?▼

Session logs may be missing if Claude Code was not launched with a named session or if Docker mounts are improperly configured. The tool includes troubleshooting guidance to help locate logs in ~/.claude/ and resolve common environment issues.

Does session debugging work for nested subagent activity and project memory layers?▼

Session debugging works for nested subagent activity by displaying nested agent trees with cost and duration metrics. It also browses project memory layers to provide full visibility into how subagents interacted during the run.