You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
[FEATURE]:Enable persistent memory for the system, supporting information retention across sessions and at the project level #32658
I'd like to work on this. My design uses SQLite + FTS5 (no new infra), post-turn extraction via a compact LLM call, and SystemContext injection as a \core/relevant-memories\ source. Phased approach: CRUD + TUI panel first, then auto-extraction, then agent tools (\search_memories\ / \store_memory). I've written a detailed design doc. Happy to open a feature issue with the full breakdown if that's preferred.
I would recommend hosting supermemory locally on your machine if you do not want to connect to any more third party services. however their hosted environment is a great starting point to learn how to interact with the service. pretty much solves the agent memory need and can be bundled up and carried across almost every llm/agent/platform you use. they also provide ample documentation and SDK's to implement the service wherever you want where they don't already have native built in plugins or you can't connect via an MCP server.
Activity
github-actions commentedon Jun 17, 2026
This issue might be a duplicate of existing issues. Please check:
Please review these issues to see if your request is already covered, and consider contributing to the existing discussions.
fixXxerTech commentedon Jul 29, 2026
I'd like to work on this. My design uses SQLite + FTS5 (no new infra), post-turn extraction via a compact LLM call, and SystemContext injection as a \core/relevant-memories\ source. Phased approach: CRUD + TUI panel first, then auto-extraction, then agent tools (\search_memories\ / \store_memory). I've written a detailed design doc. Happy to open a feature issue with the full breakdown if that's preferred.
Drewlius commentedon Aug 2, 2026
I would recommend hosting supermemory locally on your machine if you do not want to connect to any more third party services. however their hosted environment is a great starting point to learn how to interact with the service. pretty much solves the agent memory need and can be bundled up and carried across almost every llm/agent/platform you use. they also provide ample documentation and SDK's to implement the service wherever you want where they don't already have native built in plugins or you can't connect via an MCP server.
github-actions commentedon Oct 1, 2026
To stay organized issues are automatically closed after 60 days of no activity. If the issue is still relevant please open a new one.