Does this skill require external tools?

Yes, it uses a Python-based backend. A one-time setup script installs a virtual environment and the fastembed library for semantic similarity searching.

How does Memory Checkpoint prevent context loss?

It captures a detailed snapshot of your current work state, decisions, and files, allowing you to re-inject that specific context into the conversation later if the AI's context window is cleared.

What information is stored in a checkpoint?

Each checkpoint includes current status, task descriptions, context/reasoning, decisions made, pending items, blockers, next steps, and key files.

Can I search through my past checkpoints?

Yes, by providing a query after the command (e.g., '/memory-checkpoint auth bug'), the skill will perform a semantic search to find the most relevant past entries.

Memory Checkpoint

Name: Memory Checkpoint
Author: jbdamask

byjbdamask

•

Productivity & Workflow

Maintains session continuity by saving and searching structured state checkpoints to prevent context loss during long coding tasks.

The Memory Checkpoint skill solves the challenge of context drift and memory limitations in AI-assisted coding. It allows users to create structured snapshots of their current progress—including task status, reasoning, decisions made, and pending items—which are then stored and indexed. By leveraging semantic search through fastembed, users can instantly retrieve specific past states or historical context using natural language queries. This is particularly useful for complex projects where maintaining a coherent thread of development across multiple sessions or context window resets is critical for accuracy and efficiency.

Key Features

01Project-specific persistence within the local working directory

02Structured checkpointing including status, blockers, and decision history

03Semantic search for past session states using natural language queries

044 GitHub stars

05Automated setup with virtual environment and fastembed support

06Full content retrieval of top-ranked historical checkpoints

Use Cases

01Searching for the rationale behind architectural decisions made earlier in a project

02Tracking progress and pending items during multi-step refactoring workflows

03Resuming complex development tasks after a context reset or break

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jbdamask/john-claude-skills memory-checkpoint

For use in Claude.ai and ChatGPT

Download Skill