Waggle-mcp
Active·★ 44·Apache-2.0·Updated 2026-10-16
Waggle-mcp provides AI agents with a persistent, graph-backed memory that remembers decisions, reasons, and contradictions across sessions, overcoming the limitations of short context windows. It acts as a local-first memory layer for MCP-compatible AI clients, allowing for rich, structured context retrieval instead of simple text snippets.
#Persistent Memory#Knowledge Graph#AI Agent Memory#Context Management#Local-first#MCP Server#Semantic Retrieval#SQLite Backend
01
Features
01Persistent graph memory with SQLite backend by default, with optional Neo4j support.
02Integrates as an MCP server with automatic memory hooks for various AI clients like Claude, Cursor, and Gemini CLI.
03Recursive context assembly to provide compact, high-signal context packs, optimizing token usage.
04Stores relationships between facts, decisions, and contradictions, enabling temporal queries, diffs, and conflict resolution.
05Local embeddings with a deterministic offline fallback, ensuring privacy and local operation.
02
Compatibility
Python
Runtime
Verified via docs
MCP Protocol
Standard
Verified via docs
VS Code
IDE Extension
Verified via docs
Claude
AI Client
Verified via docs
Cursor
AI Client
Verified via docs
Gemini CLI
AI Client
Verified via docs
03
Quick start
1
$ pipx install waggle-mcp
2
$ waggle-mcp setup --yes
04
Use cases
↳Maintaining long-term conversational memory for AI coding assistants across development sessions.
↳Retrieving past decisions, constraints, and preferences to inform current AI agent actions.
↳Tracking changes and resolving contradictions in evolving project requirements or design choices.
↳Exporting and sharing AI agent memory snapshots for collaboration or migration between machines/sessions.
↳Debugging AI agent reasoning and understanding context by inspecting the underlying knowledge graph via Graph Studio.
05
Alternatives
MCP-memento★ 11
Local-first MCP server that gives AI agents durable repo memory and fast semantic code context.
squish-memory★ 12
Local-first memory runtime for AI agents. One command to connect Claude Code, Codex, ChatGPT, MCP workflows, and local agents.
projectmem★ 868
Local-first AI coding memory for AI agents. Records issues, attempts, fixes and decisions, then warns your agent before it repeats an approach that already failed. Native MCP server for Claude Code, Cursor, Antigravity and Codex. 100% local, no cloud, no telemetry. MIT.
mcp-local-memory★ 60
A lightweight, powerful local memory server for AI agents supporting text, entities, and relations. Enables persistent codebase understanding and user preference management.
deja-vu★ 1.3k
Your agents already solved this. deja finds it — retroactive search and MCP recall over the session logs your coding agents already wrote. Cross-agent, SSH-portable, one zero-dep binary, fully local.
Comments
Log in to leave a comment
No comments yet. Be the first!