Roadmap
What's coming next for MemNexus.
MemNexus is actively developed. Here's what we're working on and what's planned.
In progress
Documentation site
Full documentation at docs.memnexus.ai with guides, reference, AI agent content, and machine-readable llms.txt for agent discovery.
Planned
Multi-user and teams
Shared memory spaces for teams. Role-based access control, shared knowledge graphs, and team-wide pattern detection.
Memory retention policies
Automatic archival and decay for old memories. Configure retention periods per topic or memory type.
Improved entity extraction
Better automatic extraction of entities, relationships, and structured facts from memory content.
Webhook integrations
Push notifications when memories match specific patterns or topics. Useful for automated workflows and monitoring.
Additional MCP resources
Expose frequently-accessed data (recent memories, active conversations) as MCP resources for faster agent access.
Completed recently
- One-command setup (
mx setup) — configures MCP clients, skills, and hooks - Built-in MCP bridge (
mx mcp serve) with OAuth 2.1 remote clients (OpenCode, Codex, ChatGPT Desktop) - Automatic commit context capture (CommitContext)
- Automatic memory capture via end-user lifecycle hooks
- TypeScript SDK v1.53 (
@memnexus-ai/typescript-sdk) - Auto-generated reference docs — the CLI reference is derived from command source and the API reference from the OpenAPI spec, so both reflect the commands and endpoints that actually ship
- Automatic content extraction (v1.26.0)
- Prometheus monitoring and Grafana dashboards (v1.26.2)
- Narrative thread management (v1.25.0)
- Billing and subscriptions (v1.25.0)
- Behavioral pattern detection (v1.24.0)
- GraphRAG queries and entity discovery (v1.23.0)
Feature requests
Have an idea? Open an issue on GitHub with the feature-request label.