abhigyan-shekhar-waggle-mcp
Persistent graph memory for AI agents. Drop a conversation turn in via `observe_conversation()` and facts are auto-extracted, stored as typed graph nodes with local semantic embeddings (no API key). Supports temporal queries ("what did we decide last week?"), conflict detection, and context priming. One-command setup with `waggle-mcp init`. SQLite locally, Neo4j in production.
- Score
- 94.6302 signals
- Evidence
- 39 stars · 138 forks
- Last commit
- as last read from GitHub; most reads are from 2 Sep 2026 or later
- Listed
Install
No one-command install. Set it up from its source.
Alternatives · MCPs
- megamem79 stars · 12 forks89.146
- knowall-ai-neo4j-agent-memory69 stars · 14 forks88.863
- mimir68 stars · 13 forks88.594
What it is
Persistent graph memory for AI agents. Drop a conversation turn in via observe_conversation() and facts are auto-extracted, stored as typed graph nodes with local semantic embeddings (no API key). Supports temporal queries ("what did we decide last week?"), conflict detection, and context priming. One-command setup with waggle-mcp init. SQLite locally, Neo4j in production.
When to use it
When an agent needs the "Knowledge & Memory" capability this MCP server exposes.
Source
Migrated from the awesome-mcp-servers navigation directory (category: Knowledge & Memory). See https://github.com/Abhigyan-Shekhar/Waggle-mcp.