skills.rest
joySUSY
joySUSYCOMMUNITY
Explore all skills by this author →
GitHub

history-insight

Extracts and analyzes conversation insights from Claude Code session JSONL history files.

Updated Feb 21, 2026
Search Tags:#summarization#claude code#conversation analysis#jq#session history#jsonl#insight extraction
One-click install
npx skills add https://github.com/joySUSY/violet-plugin-place --skill history-insight-joysusy
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: history-insight
Source: https://github.com/joySUSY/violet-plugin-place/tree/main/plugins/developer-tool/ai-agent-memory/v3-expansion/history-insight
Command: npx skills add https://github.com/joySUSY/violet-plugin-place --skill history-insight-joysusy

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires jq, and includes scripts (resource) and references (resource) components.

What problem does it solve? Claude Code session histories accumulate as large JSONL files where 94% of the content is metadata noise, making it hard to revisit past discussions, capture decisions, or summarize prior work. ## Core Features & Use Cases - Session Discovery: Locates session JSONL files for the current project or across all projects under ~/.claude/projects, with optional date filtering. - Noise Reduction: Strips file-history snapshots, queue operations, thinking blocks, and tool calls, reducing a 12MB session file to roughly 800KB of actual conversation. - Scalable Analysis: Handles 1-3 files via direct parsing and larger sets through a batch extraction pipeline with parallel subagent summarization. - Use Case: Ask to summarize "what we discussed last week" and get a structured report of extracted insights from your session history. ## Quick Start Capture and summarize my Claude Code session history for the current project from the past week.

Frequently Asked Questions about history-insight

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

FAQPage Schema
How do I access my Claude Code session history?▼

Claude Code stores session history as JSONL files under ~/.claude/projects/<encoded-cwd>/. This skill locates those files for the current project or all projects, then parses and summarizes the conversation content.

How to extract conversation text from Claude Code JSONL files?▼

Use jq to filter JSONL entries by type, keeping user messages and assistant text blocks while discarding file-history snapshots, queue operations, thinking, and tool_use entries. The included extract-session.sh script automates this filtering.

Why are Claude Code session files so large?▼

Roughly 94% of a session file is metadata: file-history snapshots take about 67% and queue operations about 27%. Actual user and assistant conversation is only around 6%, which is why extraction shrinks files dramatically.

What happens if jq is not installed?▼

The extraction script requires jq and exits with an error if it is missing. Install it with a package manager such as brew install jq on macOS before running session extraction.

Can I analyze sessions from all projects at once?▼

Yes. Choosing the all-sessions scope searches ~/.claude/projects recursively for JSONL files. Large result sets are processed through a batch pipeline with parallel subagent summarization.

Related Skills

Explore More Software Engineering Skills→
d3mingd3ming

analyze-token-usage

Parse Claude Code JSONL session logs and aggregate token costs into markdown reports.

Community
Intermediate
aget-frameworkaget-framework
2

aget-record-lesson

Capture and classify session lessons into Framework or Domain records.

Official
Advanced
berserkdisruptorsberserkdisruptors
145

recall

Reconstruct development context from git commit history into structured intent and decision lines.

Official
Advanced
zai-orgzai-org
588

session-logs

Extract messages, timestamps, and roles from session JSONL files with jq.

Official
Intermediate