checkpointing

Persist session context and CLI interactions into agent configuration checkpoints.

Updated Feb 1, 2026
One-click install
npx skills add https://github.com/yoshihiko555/ai-orchestra --skill checkpointing-yoshihiko555
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: checkpointing
Source: https://github.com/yoshihiko555/ai-orchestra/tree/main/skills/checkpointing
Command: npx skills add https://github.com/yoshihiko555/ai-orchestra --skill checkpointing-yoshihiko555

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill automates the persistence of session context and CLI interaction history into agent configuration files and checkpoint records, enabling seamless recovery and pattern discovery across sessions.

Core Features & Use Cases

  • Session history capture: Updates Claude, Codex, and Gemini context files with a traceable history of conversations, commands, and decisions.
  • Full checkpoints: Generates comprehensive markdown checkpoints (.claude/checkpoints) that summarize Git activity, file changes, and CLI consultations for auditing and reuse.
  • Skill analysis readiness: When invoked with the analyze option, produces prompts for AI-driven pattern extraction to propose new skills.

Quick Start

Run in session history mode by default to update agent context files:

  • python checkpoint.py To create a full checkpoint and optionally analyze for skills:
  • python checkpoint.py --full
  • python checkpoint.py --full --analyze

Frequently Asked Questions about checkpointing

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

FAQPage Schema
How do I save CLI session history and git activity for AI agent context?

To save CLI session history and git activity, this skill reads local logs and git data, then writes structured checkpoints and context files for Claude, Codex, and Gemini. This ensures your AI agents retain traceable conversation and command histories across sessions.

What is the best way to capture development session context for future pattern extraction?

Capturing development session context for pattern extraction involves generating markdown checkpoints summarizing file changes and CLI consultations. When run with the analyze option, it emits an AI-analysis prompt to propose new skills based on extracted patterns.

How do I generate full checkpoints of git history and file changes for auditing?

You generate full checkpoints of git history and file changes by running the checkpoint script with the --full flag. This creates comprehensive markdown records in the .claude/checkpoints directory, summarizing your recent git activity and CLI interactions for auditing.

Can I use this checkpointing skill with multiple AI agent configuration files?

Yes, you can use this checkpointing skill with multiple AI agent configuration files. It updates context files for Claude, Codex, and Gemini by default, ensuring that session histories, commands, and decisions are traceable across different agent platforms.

Do I need any external dependencies to persist session context into checkpoints?

You do not need any external dependencies to persist session context into checkpoints. The skill operates independently using local logs and git data, writing structured checkpoints directly into the .claude/checkpoints directory without requiring additional packages.

When should I run the analyze option during session checkpointing?

You should run the analyze option during session checkpointing when you want to perform AI-driven pattern extraction on your CLI interactions and git activity. This generates a prompt designed to propose new skills based on your captured development workflow patterns.