Can I filter my memory searches by project name?

Yes, the search, timeline, and fetch tools all support a 'project' parameter to isolate memories to specific codebases or repositories.

Does this skill work across different terminal or CLI sessions?

Yes, it utilizes a persistent database to ensure that work performed and observations made in one session are searchable in any future session.

What types of coding events can this skill remember?

It can track and categorize various event types including bugfixes, features, architectural decisions, discoveries, and general code changes.

When should I use the timeline tool instead of just searching?

Use the timeline tool when you need to understand the context or sequence of events immediately before or after a specific discovery or code change.

How does mem-search save tokens during operation?

It follows a 3-layer workflow where it first retrieves low-token indexes, then filters for relevant IDs, and finally only fetches the full details for the specific items needed.

Memory Search & Persistence

Name: Memory Search & Persistence
Author: coniferconcepts

byconiferconcepts

0•

生産性とワークフロー

Enables Claude to search and retrieve context from previous coding sessions to maintain continuity and recall past solutions.

The mem-search skill provides Claude with a persistent memory layer, allowing it to remember past decisions, bug fixes, and architectural discoveries across different chat sessions. By implementing a high-efficiency three-layer workflow—Search, Timeline, and Fetch—it ensures that Claude can quickly locate specific historical data without wasting tokens on irrelevant details. This is particularly useful for long-term projects where developers need the AI to recall why certain changes were made or how specific problems were previously solved, effectively bridging the gap between disconnected development cycles.

主な機能

01Manual memory saving for architectural decisions and discoveries

02Multi-project support to isolate memory pools by codebase

030 GitHub stars

04Chronological timeline reconstruction for contextual understanding

05Token-optimized batch fetching of detailed project observations

06Persistent cross-session memory search with advanced filtering

ユースケース

01Retrieving historical project decisions when starting a new phase of a long-term project

02Recalling how a specific authentication bug was solved in a previous development sprint

03Maintaining a continuous development narrative across multiple disconnected terminal sessions

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add coniferconcepts/claude-mem-source mem-search

For use in Claude.ai and ChatGPT

Download Skill