memory-workflow

Standardize cross-session memory workflows across claude-mem, Claude-Flow, Serena, and Linear.

Updated Feb 10, 2026
One-click install
npx skills add https://github.com/monktui/git-review --skill memory-workflow-monktui
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: memory-workflow
Source: https://github.com/monktui/git-review/tree/main/.claude/skills/memory-workflow
Command: npx skills add https://github.com/monktui/git-review --skill memory-workflow-monktui

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It prevents duplicated or conflicting notes by defining where decisions, requirements, and observations must be stored across memory systems while enforcing an efficient workflow for saving and retrieving knowledge.

Core Features & Use Cases

  • Data workflow enforcement: Keeps a single source of truth by routing requirements/specs/status/acceptance criteria to Linear comments while allowing local plan files only for implementation-level detail written by the planner agent.
  • Memory system separation: Establishes clear ownership of information across claude-mem, Claude-Flow, Serena, and Linear to eliminate overlap and confusion.
  • Efficient claude-mem retrieval workflow: Requires search-to-filter-to-get-observations sequencing to avoid unnecessary token usage, and defines when to save or avoid saving across sessions.

Quick Start

Use the memory-workflow skill when you need to record a cross-session architectural decision or capture an observation for future reuse without duplicating content that already belongs in Linear.

Frequently Asked Questions about memory-workflow

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

FAQPage Schema
How to prevent duplicated memory when logging cross-session architectural decisions?

Prevent duplicated memory by enforcing a memory management workflow that separates storage ownership across claude-mem, Claude-Flow, Serena, and Linear, eliminating overlap and keeping a single source of truth for architectural decisions.

What is the best way to reduce token usage during claude-mem retrieval?

The best way to reduce token usage during claude-mem retrieval is to apply a search-to-filter-to-get_observations sequencing workflow, which restricts data fetching to only necessary observations and prevents loading unnecessary context.

How do I manage requirements tracking across Linear and local plan files?

Manage requirements tracking by routing requirements, specs, and acceptance criteria to Linear comments as the source of truth, while restricting local plan-file writing to the planner agent for implementation-level details only.

Does the memory-workflow skill integrate with Linear for project management?

Yes, the memory-workflow skill integrates with Linear by enforcing Linear comments as the source of truth for project management requirements, specs, status, and acceptance criteria during review and implementation activities.

When should I avoid saving observations across sessions?

You should avoid saving observations across sessions when the content already belongs in Linear comments or local plan files, preventing overlap and maintaining strict memory system separation between claude-mem, Claude-Flow, and Serena.

Can I use Serena and Claude-Flow together for debugging insights and knowledge workflow?

Yes, Serena and Claude-Flow can be used together by establishing clear ownership of information across memory systems, allowing you to capture debugging insights and project management requirements without overlap or confusion during implementation activities.