spfcraze

MindBank

Name: MindBank
Author: spfcraze

AI community

SOURCE https://github.com/spfcraze/MindBank

Description

MindBank gives your AI Hermes or claude code a permanent, searchable, relationship-aware memory that persists across conversations. Instead of forgetting everything between sessions, your agent remembers decisions, configs, preferences, and project knowledge — organized as a graph, not a flat text blob.

Installation

This entry records only its repository, not the path inside it, so there is no exact command to give. Open the source below and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

README

MindBank

**Persistent graph memory for AI agents — a global workspace for your fleet of sessions.**

Hybrid search · Temporal versioning · Local embeddings · Per-project isolation · MCP server · Live global-workspace (JSPACE)

[![Go](https://img.shields.io/badge/Go-1.25%2B-00ADD8?logo=go)](https://go.dev) [![Postgres](https://img.shields.io/badge/Postgres-16%2B-4169E1?logo=postgresql)](https://postgresql.org) [![Ollama](https://img.shields.io/badge/Ollama-local-000000?logo=ollama)](https://ollama.com) [![License](https://img.shields.io/badge/license-MIT-green.svg)](LICENSE)

⚡ One-Command Install

curl -sSL https://raw.githubusercontent.com/spfcraze/MindBank/main/scripts/install.sh | bash

That single command: checks prerequisites (Git, Go 1.25+, Docker), clones MindBank, starts the **pgvector** Postgres container, builds the API + MCP servers, runs migrations, and starts everything.

**Dashboard:** http://localhost:8095 · **MCP:** http://localhost:8096/mcp

Manual setup: see [QUICKSTART.md](QUICKSTART.md). Config reference: see [docs/ARCHITECTURE.md](docs/ARCHITECTURE.md) and [docs/API.md](docs/API.md).

What is MindBank?

MindBank is a **persistent, graph-structured memory layer** that sits between your AI agents and raw conversation history. Agents stop losing context between sessions: everything important they learn is written once, deduplicated, embedded, and linked into a growing knowledge graph — then recalled with hybrid full-text + semantic search.

It is also a **global workspace** (JSPACE) for many concurrent agent sessions: memories are broadcast into a shared, leased workspace, integrated into the graph, and surfaced live so the whole fleet shares what any one session learned.