skills.rest
kyletabor
kyletaborCOMMUNITY
Explore all skills by this author →
GitHub

session-historian

Analyze Claude Code JSONL session histories to diagnose bugs and trace errors.

1|Updated Feb 2, 2026
Search Tags:#context#debug#analyze#session-history#agent-behavior#workflow-analysis
One-click install
npx skills add https://github.com/kyletabor/claude_plugins --skill session-historian
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: session-historian
Source: https://github.com/kyletabor/claude_plugins/tree/main/session-historian/skills/session-historian
Command: npx skills add https://github.com/kyletabor/claude_plugins --skill session-historian

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Read and analyze Claude Code session history stored in ~/.claude/projects/. Use when asking about previous sessions, debugging past workflows, finding error patterns, searching conversations, understanding what agents did, or maintaining continuity across sessions.

Core Features & Use Cases

  • Read, summarize, and search Claude Code session histories to diagnose issues and trace agent behavior.
  • Cross-session analysis to identify recurring failures, workflow bottlenecks, and context gaps for continued work.
  • Retrieve detailed context for debugging, including messages, tool calls, and touched files.

Quick Start

Run the list_sessions.py script to list recent sessions for your project and review the summaries.

Frequently Asked Questions about session-historian

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

FAQPage Schema
How do I analyze Claude Code session histories to trace agent behavior?▼

You can trace agent behavior by running scripts that load JSONL session data from ~/.claude/projects/, generating summaries and searching past conversations to retrieve detailed context including messages, tool calls, and touched files.

What is the best way to debug past Claude Code conversations and find error patterns?▼

Debug past Claude conversations by performing cross-session analysis on JSONL history data to identify recurring failures, workflow bottlenecks, and context gaps for continued work across your projects.

Can I search across multiple Claude Code sessions to find specific tool calls or touched files?▼

Yes, you can search across sessions by running scripts that load JSONL session data to retrieve detailed debugging context, including specific messages, tool calls, and touched files across your projects.

Do I need any external dependencies to read and summarize Claude Code session history?▼

No, you do not need external dependencies. The Skill uses standalone scripts to read, summarize, and search JSONL session data directly from your ~/.claude/projects/ directory without additional packages.

How do I maintain continuity across previous Claude Code agent sessions?▼

Maintain continuity by performing cross-session analysis on JSONL history data to identify workflow bottlenecks, context gaps, and recurring failures, enabling you to resume past work effectively.

Related Skills

Explore More Software Engineering Skills→
TaylorHustonTaylorHuston
69

end-session

End active learning sessions by generating summaries and updating session logs.

Community
Basic
carlkiblercarlkibler
3

agent-log-forensics

Analyze agent session logs to detect issues and workflow bottlenecks.

Community
Intermediate
auxi-wardrobeauxi-wardrobe
1

ck:context-engineering

Analyze and optimize context usage in large language models.

Official
Advanced
alangebalangeb

error-recovery

Analyze audit logs and retry failed API calls with exponential backoff.

Community
Advanced