Is your feature request related to a problem? Please describe.
KaibanJS does team-level memory management for routing task results through workflows, but there's no built-in option for durable, cross-session long-term memory — facts and consolidated knowledge an agent team should remember across runs, not just within a board.
Describe the solution you'd like
An optional integration with Hindsight, an open-source long-term agent memory system. It organizes memory into world facts, experience facts, and consolidated mental models, with multi-strategy recall (semantic + BM25 + graph + temporal) and a reflect operation. It exposes a REST API and an MCP server.
The shape: agents recall relevant long-term memories into context before acting and retain new memories after a task completes — complementing KaibanJS's existing team-level memory rather than replacing it. I'm the maintainer and happy to build and maintain it.
Describe alternatives you've considered
- A KaibanJS tool (like the existing RAG tools) that wraps Hindsight recall/retain.
- A team-level memory adapter, if you'd prefer it integrated at that layer.
Additional context
Prior art: Hindsight already ships as a first-party memory provider in NousResearch's Hermes agent, alongside honcho, mem0, supermemory, and byterover. Mainly want to gauge interest and preferred surface before opening a PR.
Is your feature request related to a problem? Please describe.
KaibanJS does team-level memory management for routing task results through workflows, but there's no built-in option for durable, cross-session long-term memory — facts and consolidated knowledge an agent team should remember across runs, not just within a board.
Describe the solution you'd like
An optional integration with Hindsight, an open-source long-term agent memory system. It organizes memory into world facts, experience facts, and consolidated mental models, with multi-strategy recall (semantic + BM25 + graph + temporal) and a
reflectoperation. It exposes a REST API and an MCP server.The shape: agents recall relevant long-term memories into context before acting and retain new memories after a task completes — complementing KaibanJS's existing team-level memory rather than replacing it. I'm the maintainer and happy to build and maintain it.
Describe alternatives you've considered
Additional context
Prior art: Hindsight already ships as a first-party memory provider in NousResearch's Hermes agent, alongside honcho, mem0, supermemory, and byterover. Mainly want to gauge interest and preferred surface before opening a PR.