Can I use Ctx to share information between different projects?

Yes, Ctx supports project tags, allowing you to store cross-project knowledge that can be recalled regardless of which repository you are currently working in.

How does Ctx differ from the standard MEMORY.md?

While MEMORY.md is best for brief, project-scoped notes, Ctx is a structured knowledge graph that supports metadata, different data types (like hypotheses or patterns), and tiered loading to save context tokens.

Will Ctx bloat my context window usage?

No, Ctx uses a tiering system. Only 'pinned' and 'working' information is automatically loaded, while 'reference' and 'off-context' data is only retrieved when explicitly queried.

Is the memory storage automatic?

The skill includes hooks that run at the start and end of sessions, but the AI is instructed to actively emit storage commands whenever a significant decision or discovery is made.

Ctx Persistent Memory System

Name: Ctx Persistent Memory System
Author: Zate

byZate

•

开发者工具

Implements a structured, multi-tiered persistent memory system to retain decisions, patterns, and facts across Claude sessions.

The Ctx skill provides a sophisticated persistent memory layer for Claude, moving beyond basic file-based notes to a structured knowledge graph. By utilizing specialized XML-style commands, it allows Claude to categorize information into tiers—ranging from critical pinned facts to temporary working context. This ensures that architectural decisions, debugging observations, and user preferences are never lost between stateless sessions, effectively creating a long-term 'brain' for complex, multi-day development projects and cross-project knowledge management.

主要功能

01Tiered lifecycle management (Pinned, Reference, Working, Off-Context)

02Intelligent coordination and deduplication with project MEMORY.md files

036 GitHub stars

04Automatic lifecycle management via session start and end hooks

05Structured storage for decisions, patterns, facts, and hypotheses

06Advanced query and recall capabilities for on-demand knowledge retrieval

使用场景

01Tracking debugging hypotheses and root cause observations during multi-session troubleshooting

02Storing user-specific preferences and project-wide patterns to ensure consistency without repetitive prompting

03Retaining complex architectural decisions and coding conventions across a long-term project lifecycle

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add zate/cc-plugins using-ctx

For use in Claude.ai and ChatGPT

Download Skill