Add GoodMem as a Memory / Knowledge-base Node

[bashareid]

GoodMem is a memory layer for AI agents with support for semantic storage, retrieval, and summarization. GoodMem supports built-in OCR for PDF and DOCX files, so documents can be uploaded directly and turned into searchable memories without a separate ingestion pipeline. It is self-hostable, so your data stays on your own infrastructure, and it is model-agnostic — you bring your own embedder, reranker, and LLM (OpenAI, Cohere, Voyage, local models, …) and swap them per space without re-architecting. On top of that it offers multi-space semantic search (one retrieval call across many knowledge bases), full CRUD over spaces and memories with metadata, and a single API that ties embeddings, chunking, retrieval, and reranking together.

Why add it to Flowise?
Flowise already has great building blocks for chatflows and agents, but persistent, queryable memory across files/emails/web content usually means stitching together a vector DB + embedder + reranker + ingestion pipeline yourself. A GoodMem node would give Flowise users that whole stack as one node they can drop into any flow.

Key perks:

• Self-hostable — run GoodMem on your own infra, keep your data in your network.
• Bring-your-own models — choose the embedder, reranker, and LLM per space; swap them later without re-architecting.
• Built-in OCR for PDF & DOCX — upload files directly as memories, no extra ingestion node needed.
• Multi-space semantic search — one retrieval call can span many knowledge bases at once.
• Full CRUD — spaces and memories are first-class, with metadata, filters, and lifecycle management.
  Proposed Node

A single GoodMem node in the Tools category, backed by a GoodMem API credential (Base URL + API Key + Verify SSL). The node exposes 11 actions and the Default Embedder and Default Space selectors are pulled live from the connected GoodMem server, so users pick from real options instead of typing IDs. Agents can search across one or multiple spaces in a single call and manage the full lifecycle of spaces and memories from inside a Chatflow or Agentflow.

Actions: List Embedders, List Spaces, Get Space, Create Space, Update Space, Delete Space, Create Memory, List Memories, Get Memory, Retrieve Memories, Delete Memory.

Example Scenario
Build a self-hosted "ask my stuff" agent in Flowise:

Spin up GoodMem on your own server and create a space using your preferred embedder and reranker.
Ingest your emails, Google Drive files, and PDFs as memories (GoodMem OCRs PDF/DOCX on the way in).
In a Flowise Agentflow, wire the GoodMem node alongside your chosen LLM node.
The agent calls Retrieve Memories to semantically search across the relevant spaces, then the LLM summarizes the results — all running on infrastructure you control.