irisxc4/memoryguard

Overview

Local-first MCP memory governance for coding agents including Codex, Claude Code, Cursor, Grok, and Trae. It provides shared rules, deduplication, token insights, audit, and rollback for agent memory. Built in Python and runs locally.

Best for

Best for
Developers using multiple AI coding agents who want local memory governance with audit and rollback

Use cases

• Centralize shared memory rules across multiple coding agents
• Deduplicate redundant memory entries to reduce token usage
• Audit and rollback memory changes for debugging or recovery

How to use

Install

python -m pip install agent-memguard

Tools exposed

• mcp-status
• memoryguard_context_bootstrap
• memoryguard_memory_search
• memoryguard_memory_read
• memoryguard_memory_write
• memoryguard_memory_update
• memoryguard_memory_delete
• memoryguard_memory_status
• memoryguard_audit
• memoryguard_explain

Tested with

Claude Code, Cursor

Notes

Local-first MCP memory governance for coding agents including Codex, Claude Code, Cursor, Grok, and Trae. It provides shared rules, deduplication, token insights, audit, and rollback for agent memory. Built in Python and runs locally.

3 stars on GitHub. Last updated 2026-09-05. Licensed MIT.

Use cases

• Centralize shared memory rules across multiple coding agents
• Deduplicate redundant memory entries to reduce token usage
• Audit and rollback memory changes for debugging or recovery

Pros

• Local-first design keeps memory data on your machine
• Supports several popular coding agents
• Includes audit and rollback for safer memory management

Cons

• Very low community adoption with only 3 stars
• Requires a Python environment to run
• Limited track record for reliability or maintenance

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.