transcript-condenser

Condenses Claude Code session transcripts from .claude/logs into Markdown summaries and JSON reports.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/OrasesLabs/orases-claude-code-marketplace --skill transcript-condenser-oraseslabs
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: transcript-condenser
Source: https://github.com/OrasesLabs/orases-claude-code-marketplace/tree/main/workflow-analyzer/skills/transcript-condenser
Command: npx skills add https://github.com/OrasesLabs/orases-claude-code-marketplace --skill transcript-condenser-oraseslabs

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill condenses verbose Claude Code session transcripts into readable summaries by filtering metadata, system notifications, and command artifacts, enabling quicker review and sharing of what happened during a session.

Core Features & Use Cases

  • Output formats: Markdown summaries and machine-readable JSON for automation
  • Verbosity control: Minimal, standard, and detailed views for quick reviews or deep analyses
  • Batch processing: Process directories of transcripts with consistent output for reports

Quick Start

Condense a transcript to a Markdown summary or export JSON for downstream processing, using your transcript file as input.

Frequently Asked Questions about transcript-condenser

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

FAQPage Schema
How do I condense Claude Code session transcripts into summaries?

Transcript condensing filters verbose Claude Code logs into readable summaries by removing metadata, system notifications, and command artifacts. Point the Skill at your .claude/logs/ transcript file to generate Markdown or JSON output in minimal, standard, or detailed verbosity.

What output formats does transcript summarization support?

Transcript summarization produces Markdown summaries for human review and machine-readable JSON for automation and downstream processing. Both formats preserve session context while stripping noise for faster analysis.

Can I filter transcript summaries to show only specific session activity?

Yes, filtering options let you isolate tool usage, subagent interactions, or strip system messages from summaries. Configurable verbosity and filtering enable focused review of particular session aspects.

How do I process multiple session transcripts at once?

Batch processing lets you condense directories of transcripts with consistent output formatting for generating reports, comparing session patterns, and exporting analysis-ready data across multiple sessions.

What's the best way to review and share Claude Code session history?

Condense transcripts into readable summaries to quickly review what happened, identify tool usage patterns, and share session reports without exposing raw logs. Markdown and JSON exports suit both human sharing and machine analysis.

Do I need to manually edit transcripts before summarization?

No. The Skill automatically filters metadata and system artifacts from .claude/logs/ transcripts. Input transcripts are processed as-is; no manual preprocessing is required.