memory-sync

Commit and push memory updates to refresh MCP documents.

Updated May 14, 2024
One-click install
npx skills add https://github.com/wagnerra23/oimpresso.com --skill memory-sync-wagnerra23
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: memory-sync
Source: https://github.com/wagnerra23/oimpresso.com/tree/main/.claude/skills/memory-sync
Command: npx skills add https://github.com/wagnerra23/oimpresso.com --skill memory-sync-wagnerra23

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The Skill ensures knowledge stored in this repo’s memory is actually visible to the team through the MCP server, preventing “it wasn’t planned” situations caused by forgetting to push changes.

Core Features & Use Cases

  • Git → MCP synchronization discipline: commits and pushes only the relevant memory updates, then relies on MCP webhook/tooling to make the team’s searches reflect the latest SPEC/ADR/session logs.
  • Correct tool-flow for tasks: directs task operations to MCP tools (tasks-* / cycles-active / cycle goals tracking) instead of Markdown-based CURRENT/TASKS files.
  • MCP-first handoff safety (append-only): enforces the required MCP snapshot checklist before creating new handoff entries in memory/handoffs to avoid accidental overwrites and information loss.

Quick Start

Use the memory-sync skill to sync any newly updated files under memory/ by running /sync-mem and pushing the required commit so the MCP team can immediately find the updates via MCP tools.

Frequently Asked Questions about memory-sync

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

FAQPage Schema
How do I sync local memory updates with the MCP server for team coordination?

To sync memory updates with the MCP server, run the /sync-mem command to detect pending changes, then commit and push only the relevant memory paths so the GitHub webhook refreshes MCP knowledge for team discovery.

Why are my architecture decisions and session logs not visible to the team via MCP tools?

Architecture decisions and session logs remain invisible when local memory updates are not committed and pushed, preventing the GitHub webhook from refreshing the MCP server to reflect the latest knowledge changes.

What is the correct workflow for creating MCP handoff entries without losing data?

Creating MCP handoff entries requires an append-only approach that enforces the MCP snapshot checklist before adding new entries to memory/handoffs, which avoids accidental overwrites and information loss.

How do I track task progress using MCP tools instead of markdown files?

Track task progress by directing task operations to MCP tools like tasks-* and cycles-active instead of markdown-based CURRENT or TASKS files, ensuring the team sees real-time cycle goals tracking.

Do I need to run a specific command before pushing memory updates to GitHub?

Yes, you must run /sync-mem for pending detection before pushing memory updates to GitHub, ensuring only the correct memory paths under memory/decisions, memory/sessions, and memory/handoffs are committed.

What limitations exist when committing memory updates for MCP knowledge synchronization?

A key limitation is that memory synchronization requires committing only the correct memory paths to trigger the GitHub webhook; pushing unrelated changes prevents the MCP server from accurately refreshing knowledge.