velixar_store
Store important facts, decisions, user preferences, and project context for long-term retrieval. Memories persist across sessions with automatic chunking and duplicate detection.
Instructions
Store a memory for later retrieval. Use for important facts, decisions, user preferences, project context, or anything worth remembering long-term. Memories are workspace-scoped and persist across sessions. Long content is automatically chunked; max 50,000 characters (larger is rejected with a 400 before embedding — split it or chunk client-side). DEDUP (before you call): exact byte-identical duplicates are always blocked server-side. If this content may overlap semantically with existing memories, either velixar_search first or pass check_duplicate:true to run a near-duplicate check.
Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| tags | No | Optional tags for categorization | |
| tier | No | Memory tier: 0=pinned, 1=session, 2=semantic (default), 3=org | |
| atomic | No | Store as a single unit — never split into chunks. Use for poems, quotes, structured data that must stay whole. | |
| source | No | Provenance label (e.g., "user-stated", "derived-from-analysis") | |
| content | Yes | The memory content to store (max 50,000 chars; longer is rejected before embedding). Long content is chunked automatically. | |
| source_ids | No | DERIVATION lineage — the memory IDs this new memory was reasoned/built FROM (e.g. you learned context B from context A). An explicit, directed "derived-from" edge, DISTINCT from semantic similarity and from the temporal previous-memory chain. Omit for an origin memory (learned fresh, no prior context). Query the resulting graph with velixar_lineage. | |
| check_duplicate | No | Exact byte-identical duplicates are ALWAYS blocked server-side regardless of this flag (existing id returned, nothing written). Set true to ADDITIONALLY run a near-duplicate similarity check — advisory: near-dupes above dedup_threshold warn but still store. | |
| dedup_threshold | No | Similarity threshold for duplicate detection (default: 0.95). Only used when check_duplicate is true. | |
| quarantine_zone | No | Optional quarantine zone ID. Memory will only be visible to zone members. |