os-save

Save session decisions and context into timestamped project memory snapshots.

4|Updated Apr 26, 2026
One-click install
npx skills add https://github.com/simonconway1979/os-intelligence --skill os-save
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: os-save
Source: https://github.com/simonconway1979/os-intelligence/tree/main/.claude/skills/os-save
Command: npx skills add https://github.com/simonconway1979/os-intelligence --skill os-save

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It eliminates lost context between AI sessions by saving what happened, extracting the decisions and current state, and turning them into a snapshot that can be retrieved later.

Core Features & Use Cases

  • Session context snapshots: Creates a timestamped memory file in the active project’s memory folder so future sessions start from an accurate, searchable baseline.
  • Decision + state capture: Updates the project’s current-state.md and sessions indexes with active work streams, open questions, and key decisions to keep the narrative coherent across time.
  • Portfolio-aware memory: Correctly routes saving for both whole-project sessions and portfolio-item sessions by following projects.md, CLAUDE.md, and TRACKER.md mappings.
  • Optional exploratory reasoning logs: When a session is exploratory, writes a dedicated reasoning appendix and links it from the main snapshot.

Quick Start

Run os-save at the end of your working session to persist the session summary, decisions, and updated project state into the correct project memory.

Frequently Asked Questions about os-save

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

FAQPage Schema
How do I save Claude Code session context and decisions before switching projects?

To save Claude Code session context, run os-save at the end of your work to extract decisions and work streams into a durable project memory snapshot for future retrieval. It updates memory files and indexes to preserve the narrative.

What is the best way to persist AI session memory and avoid lost context between sessions?

Persisting AI session memory requires creating a timestamped snapshot that captures current state and decisions. This approach writes a memory file and updates sessions indexes, ensuring future sessions start from an accurate, searchable baseline without losing context.

Can I track project decisions and work streams for multiple portfolio items?

Yes, you can track project decisions for multiple portfolio items. Portfolio-aware memory routing correctly saves both whole-project sessions and portfolio-item sessions by following projects.md, CLAUDE.md, and TRACKER.md metadata mappings.

How do I create an exploratory reasoning log for my AI working session?

To create an exploratory reasoning log, run a save command on an exploratory session. The process writes a dedicated reasoning appendix and links it from the main snapshot, preserving the exploratory narrative alongside core decisions.

Does session saving automatically commit symlinked skill changes to git?

Yes, session saving ensures git companion commits for symlinked skill changes. When the snapshot is created, it automatically handles the version control of linked skill modifications alongside updating the current-state synthesis.

When do I need to generate a project memory snapshot for AI context management?

You need to generate a project memory snapshot at the end of a working session or before switching projects. This captures active work streams and open questions, preventing lost context and ensuring future AI sessions start from an accurate baseline.