Local-first vector memory with a 4-layer associative graph. Your agent remembers why things are connected, when things changed, and which memories matter most.
Conversation turns, tool outputs, observations. Any unstructured agent context fed in as text.
Every memory is evaluated: ADD new info, UPDATE existing, DELETE contradictions, or NO_OP if already known. Zero duplicates.
Persisted across 4 graph types in SQLite. Survives all session resets. REM cycle compresses while idle.
Similarity between memories. Finds related concepts across your full context history.
Cause → Effect relationships. Understands why things happened, not just what.
Before → After sequences. Tracks how knowledge evolves and decays over time.
Named entity co-occurrence. Connects people, projects, and events automatically.
Most vector stores are passive. They store what you put in and return what you ask for. VEKTOR is an active memory layer — it evolves, curates, and reasons about what your agent should remember.
Standard RAG vector stores
VEKTOR Memory
Drop-in memory layer for LangChain agents. recall() returns context, remember() stores every turn. v1 + v2 adapters included.
Wrap your OpenAI agent loop with persistent memory. Inject recalled context directly into system prompt. GPT-4o and o-series models supported.
Full MCP server module — vektor_recall, vektor_store, vektor_graph, vektor_delta tools. Connect Claude Desktop to persistent memory in minutes.
Provider-agnostic. Pass gemini, openai, groq, or ollama as provider. Key pooling for Gemini — waterfall rotation across up to 9 API keys, zero rate-limit downtime.
vektor_memoire HTTP tool — works with Le Chat and Mistral API agents. HMAC-bound licence verification. Run mistral-setup.js to activate in 60 seconds. French-first sovereign memory.
// 1. Install
// npm install -g ./vektor-slipstream-v1.6.3.tgz
import { createMemory } from 'vektor-slipstream';
// 2. Initialise
const memory = await createMemory({
provider: 'gemini',
apiKey: process.env.GEMINI_API_KEY,
agentId: 'my-agent',
dbPath: './my-agent.db',
});
// 3. Remember — AUDN decides ADD/UPDATE/DELETE
await memory.remember("User prefers TypeScript");
// 4. Recall
const ctx = await memory.recall("coding preferences");
// 5. Traverse the graph
const g = await memory.graph("TypeScript", { hops: 2 });
// 6. What changed in 7 days?
const d = await memory.delta("architecture", 7);Pure SQLite. No cloud dependency, no API keys for memory. Your memory graph never leaves your server. LLM providers process queries per their own privacy policies.
Gemini, OpenAI, Groq, Ollama. Switch provider with one config change. Key pooling for Gemini — waterfall rotation across up to 9 keys.
Automatic curation loop prevents contradictions and duplicates. The graph stays consistent without any manual management.
Background process compresses 50 fragments into 3 core insights. Runs while your agent is idle. Run via vektor rem from the CLI.
In modern Agentic Engineering, memory is the bridge between reasoning and action. Without Persistent AI Memory, agents suffer from "Context Window Bankruptcy." VEKTOR solves this via Agentic RAG, allowing models to store causal relationships and temporal sequences in a local Vector Database structure.
Vektor is MCP-native. Our infrastructure allows agents using Claude Code or custom Agentic Frameworks to recall long-term context instantly. By utilizing the Model Context Protocol, Vektor grants agents "Object Permanence" across distinct sessions and multi-agent workflows.
As search interest in Sovereign AI reaches breakout levels, the need for local-first infrastructure is critical. VEKTOR provides Zero-Knowledge Memory, ensuring proprietary data remains on-premises while maintaining the high-speed recall required for Agentic Commerce and engineering.
Memory is what an agent knows. Identity is who it is and what it can access. Cloak enforces this separation by design — your SQLite memory graph and your encrypted passport vault are architecturally isolated. Share your memory backup freely. Your credentials stay locked to your machine.
| Feature | Slipstream | Cloud RAG |
|---|---|---|
| Recall latency | ~8ms | 200–800ms |
| Embedding cost | $0 · local | $0.0001/token |
| Data ownership | Your machine | Their servers |
| Setup | 1 curl command | SDK + API key + billing |
| Offline capable | Yes | No |
| Scales to | Millions of vectors | Depends on plan |
VEKTOR connects to Mistral via a hardware-bound HTTP tool endpoint. Your agent calls vektor_memoire directly — no local server, no MCP daemon. Memory lives on your VPS, credentials never leave your machine.
{
"function": {
"name": "vektor_memoire",
"description": "Query VEKTOR sovereign memory graph.
Returns ranked memory fragments with
importance scores.",
"parameters": {
"query": { "type": "string" },
"key": { "type": "string" },
"signature": { "type": "string" },
"limit": { "type": "integer", "default": 5 }
},
"required": ["query", "key", "signature"]
}
}// Paste into Mistral / Le Chat agent
Tu es un assistant avec accès à une
mémoire persistante via vektor_memoire.
Utilise cet outil pour récupérer le
contexte pertinent avant de répondre.
Rappelle toujours avec query = sujet
principal de la question utilisateur.
// English version also supported
You have persistent memory via the
vektor_memoire tool. Always recall
before responding to any query.npm install in the directory.node mistral-setup.js. Enter your licence key. Bridge activates in 60 seconds.vektor-tool-manifest.json as a tool in your Mistral agent or La Plateforme project.HMAC(secret, key:unix_minute). Signatures expire every 60 seconds — replay attacks are architecturally impossible. Your licence key never travels without a valid time-bound signature. The bridge validates against your Polar licence — revoked on refund, bound to your machine.
Questions? [email protected]