Skip to main content
Glama
Sign In
deenesjakoruzh
Kronvex

Kronvex

by Daftgoldens
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Python
Hybrid

Kronvex — EU-Native Memory API for AI Agents

Persistent, semantically searchable memory.
Three endpoints. GDPR-compliant. Data stays in Europe.

PyPI npm License: MIT EU Frankfurt Uptime

Why Kronvex?

Every time a user opens a new session with your AI agent, it starts from scratch. No context, no history, no user preferences. You end up injecting entire conversation histories into every prompt — expensive, slow, and context-window-limited.

Kronvex gives your agent persistent, semantically searchable memory across sessions. Store interactions, recall relevant context by meaning, inject a ready-to-use context block before each LLM call — and keep all data in Europe.

Related MCP server: Mnemoverse Memory

Performance

Endpoint

p50

p99

/remember

<30ms

<180ms

/recall

<45ms

<280ms

/inject-context

<55ms

<320ms

99.9% uptime · EU Frankfurt · GDPR-compliant · pgvector cosine similarity · 1536-dim embeddings

Quick Start

1. Get a free API key

curl -X POST https://api.kronvex.io/auth/demo \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Alice",
    "email": "alice@company.com",
    "usecase": "Customer support bot with memory"
  }'
{
  "full_key": "kv-xxxxxxxxxxxxxxxx",
  "agent_id": "uuid-of-your-first-agent",
  "memory_limit": 100,
  "message": "Ready! Your API key and first agent are set up."
}

2. Store a memory

curl -X POST https://api.kronvex.io/api/v1/agents/{agent_id}/remember \
  -H "X-API-Key: kv-xxxxxxxxxxxxxxxx" \
  -H "Content-Type: application/json" \
  -d '{"content": "Alice is a Premium customer since January 2023."}'

3. Inject context before each LLM call

curl -X POST https://api.kronvex.io/api/v1/agents/{agent_id}/inject-context \
  -H "X-API-Key: kv-xxxxxxxxxxxxxxxx" \
  -H "Content-Type: application/json" \
  -d '{"message": "I still have that billing issue"}'
{
  "context_block": "[KRONVEX CONTEXT]\n- Alice is a Premium customer since Jan 2023 (similarity: 0.94)",
  "memories_used": 1
}

SDKs

Python

pip install kronvex
from kronvex import Kronvex

kx = Kronvex("kv-your-api-key")
agent = kx.agent("your-agent-id")

await agent.remember("User prefers concise answers")
context = await agent.inject_context("How should I format this?")

Node.js / TypeScript

npm install kronvex
import { Kronvex } from "kronvex";

const kx = new Kronvex("kv-your-api-key");
const agent = kx.agent("your-agent-id");

await agent.remember("User prefers concise answers");
const context = await agent.injectContext("How should I format this?");

MCP (Claude Desktop)

{
  "mcpServers": {
    "kronvex": {
      "command": "npx",
      "args": ["kronvex-mcp"],
      "env": { "KRONVEX_API_KEY": "kv-your-api-key" }
    }
  }
}

→ Python SDK on PyPI · Node SDK on npm

How It Works

Memories are ranked by a composite confidence score:

confidence = similarity × 0.6 + recency × 0.2 + frequency × 0.2

  • Similarity: pgvector cosine similarity on 1536-dim OpenAI embeddings

  • Recency: sigmoid with 30-day inflection point

  • Frequency: log-scaled access count

Self-Hosting

# Requires Docker
cp .env.example .env
# Edit .env with your OPENAI_API_KEY and DATABASE_URL
docker-compose up --build

API available at http://localhost:8000 · Docs at http://localhost:8000/docs

Endpoints

Method

Endpoint

Description

POST

/auth/demo

Get a free API key

POST

/api/v1/agents

Create an agent

GET

/api/v1/agents

List your agents

POST

/api/v1/agents/{id}/remember

Store a memory

POST

/api/v1/agents/{id}/recall

Semantic search over memories

POST

/api/v1/agents/{id}/inject-context

Get context block

DELETE

/api/v1/agents/{id}/memories/{mid}

Delete a memory

GET

/health

Health check

Full interactive docs: api.kronvex.io/docs

Pricing

Plan

Price

Agents

Memories

Free

Free

1

100

Builder

€29/mo

5

20,000

Startup

€99/mo

15

75,000

Business

€349/mo

50

500,000

Enterprise

Custom

Unlimited

Unlimited

→ See full pricing

Contributing

See CONTRIBUTING.md.

Built in Paris · kronvex.io · hello@kronvex.io

Available Tools

4 tools
forgetA

Search for memories matching a query and permanently delete the top matches. Use this to remove outdated, incorrect, or superseded information from memory.

When to call: when the user explicitly asks to forget something, or when you detect that a stored memory is no longer accurate (e.g. a dependency was upgraded, a decision was reversed, a team member left).

Internally performs a high-threshold semantic search (0.7) to find close matches, then deletes up to 3 results. Returns a list of deleted memories or a message if nothing matched.

Inputs: query (required).

inject_contextA

Retrieve the most relevant memories for the current task and return them as a single formatted context block, ready to prepend to a prompt or include in a system message.

When to call: at the start of a complex or multi-step task where relevant project history, constraints, or preferences may exist in memory. Prefer this over recall when you want a single ready-to-use block rather than a list of individual memories.

Returns a formatted text block summarising the relevant memories, and a count of how many memories were used. Returns empty if none are relevant.

Inputs: query (required).

recallA

Search long-term memory using semantic similarity and return the most relevant stored memories ranked by a confidence score (weighted combination of similarity, recency, and access frequency).

When to call: before starting a task, when the user references something from a past session, or when you need project-specific context.

Returns a ranked list of memories with their confidence score and type. Returns an empty result if no memories exceed the similarity threshold.

Inputs: query (required), top_k.

rememberA

Persist a piece of information to long-term memory so it can be recalled in future sessions. Use this whenever the user states a preference, makes a decision, or shares context that should survive beyond the current conversation.

When to call: after learning the user's tech stack, coding conventions, project constraints, architectural decisions, or personal preferences.

Returns a confirmation message with the stored content preview.

Examples of good memories:

  • 'User prefers TypeScript strict mode with no implicit any'

  • 'Database: PostgreSQL 16 with pgvector on Railway, connection via asyncpg'

  • 'Never use any() type in this codebase — team policy'

  • 'Deployed on 2024-03-15: migrated auth from JWT to Supabase sessions'

Inputs: content (required), memory_type.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Deploy Server
    One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

    TDQS

    A4.5/5.0

    Scored across 4 tools

    Disambiguation5/5

    Each tool has a clearly distinct purpose with no overlap: forget deletes memories, inject_context formats them for immediate use, recall retrieves them as a ranked list, and remember stores new memories. The descriptions explicitly differentiate their roles, making misselection unlikely.

    Naming Consistency5/5

    All tool names follow a consistent verb-based pattern (forget, inject_context, recall, remember) that directly reflects their core actions. The naming is uniform and predictable, with no deviations in style or convention.

    Tool Count5/5

    With 4 tools, this server is well-scoped for its memory management domain, covering the essential CRUD-like operations (create, read, delete) plus a specialized formatting tool. Each tool earns its place without feeling excessive or insufficient.

    Completeness4/5

    The toolset provides strong coverage for memory operations: remember (create), recall (read), forget (delete), and inject_context (format). A minor gap exists in update functionality for modifying existing memories, but agents can work around this by deleting and re-adding.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Resources

    Related MCP Connectors

    Related MCP Servers

    • LoreConvo

      LoreConvoofficial

      labyrinth-analytics
      A
      license
      A
      quality
      B
      maintenance
      Persistent memory for AI agents across Claude Code, Cowork, Codex, Cursor and Hermes Agent. Saves, searches and recalls sessions from a local SQLite store through 40+ MCP tools, with auto-save and auto-load hooks, project tagging and export/import. Also a Hermes Agent memory provider plugin in the official catalog. Free: keyword search, 50 sessions. Pro ($8/mo): semantic search and team memory.
      33
      197 PyPI
      12
      Business Source 1.1

    • MCP Memory

      TWFBusiness
      F
      license
      Not graded
      quality
      D
      maintenance
      Persistent memory server for AI assistants with semantic search and three-layer context (global, project, personality). Works with MCP-compatible AI tools like Claude Code, Cursor, Continue, Cline, and more.
      1
      -