Abhigyan-Shekhar/Waggle-mcp

Overview

Waggle-mcp is an MCP server that provides an external memory layer for AI agents. It is installed via PyPI and allows agents to persist and retrieve information across sessions.

Best for

Best for
Developers experimenting with persistent memory for AI agents in Python

Use cases

• Give AI agents long-term memory for ongoing conversations
• Store and recall context across different agent sessions
• Build persistent knowledge bases for agent workflows

How to use

Install

pip install -e ".[dev]"

Tools exposed

• WAGGLE_RECURSIVE_CONTEXT_ENABLED
• WAGGLE_RECURSIVE_CONTEXT_DEFAULT_BUDGET
• WAGGLE_RECURSIVE_CONTEXT_MAX_SUBQUERIES
• WAGGLE_RECURSIVE_CONTEXT_DEFAULT_DEPTH
• WAGGLE_RECURSIVE_CONTEXT_INCLUDE_EVIDENCE
• observe_conversation
• query_graph
• prime_context
• graph_diff
• aggregate_graph
• get_related
• get_node_history
• get_topics
• list_conflicts
• resolve_conflict
• update_node
• delete_node
• decompose_and_store
• dedup_candidates
• canonicalize_node

Tested with

Claude Desktop, Claude Code, Cursor, Continue, VS Code, ChatGPT

Notes

Waggle-mcp is an MCP server that provides an external memory layer for AI agents. It is installed via PyPI and allows agents to persist and retrieve information across sessions.

12 stars on GitHub. Last updated 2026-06-01. Licensed Apache-2.0.

Use cases

• Give AI agents long-term memory for ongoing conversations
• Store and recall context across different agent sessions
• Build persistent knowledge bases for agent workflows

Pros

• Simple installation via pip
• Enables persistent memory for stateless agents
• Open-source and extensible

Cons

• Very early stage with only 12 GitHub stars
• Limited documentation and community support
• Unclear performance and reliability at scale

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