session-analyzer

Analyze Claude Code JSONL session logs to rank candidate skills for SKILL.md drafting.

Updated Apr 18, 2026
One-click install
npx skills add https://github.com/mkoziy/skills --skill session-analyzer-mkoziy
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: session-analyzer
Source: https://github.com/mkoziy/skills/tree/main/.claude/skills/skill-workshop
Command: npx skills add https://github.com/mkoziy/skills --skill session-analyzer-mkoziy

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill analyzes Claude Code session JSONL files to identify patterns across sessions, enabling rapid generation of new skill candidates.

Core Features & Use Cases

  • Extract and analyze multiple sessions to surface recurring patterns.
  • Score and rank candidate skills for drafting new SKILL.md entries.
  • Generate draft SKILL.md content from evidence for review.

Quick Start

Run the analysis to extract candidate skills from your session history and generate draft SKILL.md entries.

Frequently Asked Questions about session-analyzer

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

FAQPage Schema
How do I extract skill candidates from JSONL session history?

To extract skill candidates from JSONL session history, you analyze Claude Code session logs to identify repeating explanations and tool-usage patterns, which are then scored and ranked for drafting new SKILL.md entries.

What is the best way to detect repeating patterns in Claude Code sessions?

Detecting repeating patterns in Claude Code sessions involves parsing JSONL files to surface recurring workarounds and tool-usage sequences, generating structured results with scored candidate skills for review.

Can I generate draft SKILL.md entries automatically from existing logs?

You can generate draft SKILL.md entries automatically from existing logs by analyzing session histories for evidence of recurring patterns, producing draft content stored in /tmp/skill-workshop-results.json for review.

Does the session analyzer require any external dependencies to parse JSONL files?

The session analyzer requires no external dependencies to parse JSONL files, operating independently to extract evidence across multiple project sessions and score candidate skills without additional setup.

How are candidate skills ranked when analyzing multiple project sessions?

Candidate skills are ranked by score when analyzing multiple project sessions, parsing existing logs to extract cross-session evidence and returning structured results ordered by pattern significance.

What limitations exist when surfacing skill candidates from session history?

When surfacing skill candidates from session history, limitations include relying entirely on existing JSONL log quality and completeness, with outputs stored to a temporary file for manual review before drafting.