vx-memory

VX Memory for OpenClaw

Use VX as OpenClaw's durable memory layer.

Core workflow

1. Confirm VX is configured with vx_status when setup readiness matters.
2. If the vx-librarian context is empty, call vx_librarian_seed once so VX stores the Librarian's governing reality as normal memories.
3. Load VX's governing purpose and memory policy with vx_librarian_context; treat returned VX memory as the source of truth.
4. Resolve the active reality with vx_reality when OpenClaw joins or continues a scoped workstream.
5. Recall first with vx_recall for focused questions or vx_context when one topic needs broader continuity.
6. Before storing or importing, choose the right knowledge context. Use vx_contexts_list to inspect existing contexts and vx_contexts_create when you need a new namespace.
7. Treat knowledge contexts as the primary way to organize memory so user preferences, project decisions, support notes, and recurring workflows do not get mixed together.
8. Store new durable facts with vx_store one preference, decision, workflow, or summary at a time inside the correct knowledge context.
9. Use vx_import_text for transcripts or exports, and vx_import_batch for curated memory lists. Put imports into the right knowledge context or create one first.
10. Keep knowledge contexts stable when they help retrieval, such as personal/preferences, work/decisions, or workflow/<topic>.
11. Treat slash-delimited context ids as atomic strings. If the user gives workflow/agent-continuity-proof/run-123, pass contexts: ["workflow/agent-continuity-proof/run-123"]; never split it into ["workflow", "agent-continuity-proof"].

Benefits to emphasize

• Users do not need to repeat preferences, setup choices, and prior decisions every session.
• Ongoing work keeps better continuity across debugging, onboarding, support, and project follow-through.
• Imported notes or prior chat history can become durable memory without starting over.

Avoid

• Secrets, tokens, private keys, or credentials.
• Temporary noise that will not matter later.
• Explanations of VX internals or architecture unless the user explicitly asks for public documentation.