Qdrant Memory

Qdrant Memory MCP

One Memory. Multiple Agents. No Lock-in.

The Codex plugin candidate adds guided local setup, read-only native-memory import and verified JSON/Markdown handoff. It is versioned independently and does not add a second Runtime. See the executed validation and limits. Exports remain readable without this plugin or a subscription; native adoption is verified separately from file creation.

Durable, portable, and permission-aware memory for AI agents — distributed as an MCP adapter over the Qdrant Memory Runtime.

For shared production deployments, read the deployment security guide. Owner-only database write credentials are required independently of MCP grants. Configure destination protection and shared doctor before deployment. See the Owner support record for support limits.

One Memory, Multiple Agents — opt-in shared owner

The distribution adds an opt-in single-owner service and stdio bridge. Install Runtime 0.6.0 to use them. See shared integration guide. Private defaults are unchanged. Native Provider automatic shared extraction is deferred.

A WorkBuddy connector and a DeepSeek Harness bundle reuse this stdio MCP, the managed launcher and the canonical Skill.

简体中文

Persistent, scoped memory for AI agents through four explicit stdio MCP tools. This repository distributes the existing Python Runtime. Its metadata-only wheel contains no Python business implementation. The console script points directly to qdrant_memory.standalone:main and depends on hermes-plugin-qdrant-memory[mcp]==0.6.0.

See 0.3.1 release notes for the Runtime pin and upgrade guidance.

Before using an exact-version installation command, confirm that version on the PyPI release page. If unavailable, these package commands cannot run yet. Use the published 0.3.0 guide until publication; a source candidate does not establish PyPI or npm availability.

Versions: MCP distribution 0.3.1 pins Runtime 0.6.0, which supports MCP Python SDK 1.x and 2.x (mcp>=1.30,<3). The four scoped tools and stdio launch arguments are the same for both SDK majors. The commands below use these exact versions. Release artifacts are available through the distribution PyPI page and Runtime PyPI page. See release checklist for the publishing sequence.

Runtime, optional Agent Skill and MCP distribution versions are independent. The Skill is usage guidance, not a runtime dependency. Hermes retains its native MemoryProvider, automatic extraction and prefetch. Ordinary MCP performs only explicitly requested memory operations. It does not automatically extract every conversation turn.

Harness support matrix

Verification levels: configuration example only, protocol-level tested, official SDK/CLI tested, real host tested, real desktop GUI tested, not verified.

This distribution does not auto-extract conversations for any MCP host. Automatic capture and recall require a native provider or a lifecycle adapter. See the Codex validation record for the tested CLI, source revisions and model providers; desktop GUI and native-memory adoption are separate verification boundaries.

Install and initialize

For the shortest currently supported local setup, follow the Ollama quickstart (中文). It covers an existing model, installing a missing model, a separate WorkBuddy profile, connection settings and a first save/search check. No Qdrant server, Docker, cloud account or embedding API key is needed for that local path. Python/uv and one-time profile configuration are still required for this direct route. The managed WorkBuddy connector and DSH bundle prepare those dependencies from Node.js instead.

Requires Python 3.11+ and either uv or a Python virtual environment. You must configure storage and an embedding service. For the example below, run Ollama and install qwen3-embedding:4b yourself.

Create an operator-owned profile directory and save qdrant-memory.json inside:

{

"schema_version": 1, "scope": {"user_id": "my-user", "agent_id": "my-agent"}, "qdrant": {"mode": "embedded", "collection": "qdrant_memory"}, "embedding": { "mode": "plugin", "provider": "ollama", "model": "qwen3-embedding:4b", "base_url": "http://127.0.0.1:11434", "dimensions": 2560, "distance": "Cosine" } }

Initialize, then launch (replace /path/to/profile with your absolute path):

uvx --from qdrant-memory-mcp==0.3.1 qdrant-memory-mcp --home /path/to/profile init

uvx --from qdrant-memory-mcp==0.3.1 qdrant-memory-mcp --home /path/to/profile mcp

The command without --home uses ~/.qdrant-memory, independent of Hermes. Configure and initialize that directory before letting a Registry client launch uvx --from qdrant-memory-mcp==0.3.1 qdrant-memory-mcp mcp. The executable name matches the package name. The Registry's positional mcp selects protocol mode. Its optional named --home precedes mcp when a client supplies a custom profile.

Alternatively:

python3 -m venv .venv

.venv/bin/python -m pip install qdrant-memory-mcp==0.3.1 .venv/bin/qdrant-memory-mcp --home /path/to/profile init .venv/bin/qdrant-memory-mcp --home /path/to/profile mcp

On Windows use .venv\Scripts\python.exe and .venv\Scripts\qdrant-memory-mcp.exe.

Qdrant embedded, self-hosted server and Cloud, plus Ollama and OpenAI-compatible embeddings, are configured in the Runtime's configuration reference. Keep credentials in operator environment variables (QDRANT_API_KEY, EMBEDDING_API_KEY, or configured names). Never commit their values. Initialization against an existing collection requires explicit operator choice. Do not use a production destination for a first test.

Connect a client

Codex config.toml example:

[mcp_servers.qdrant_memory]

command = "uvx" args = ["--from", "qdrant-memory-mcp==0.3.1", "qdrant-memory-mcp", "--home", "/path/to/profile", "mcp"]

OpenCode opencode.json example:

{

"$schema": "https://opencode.ai/config.json", "mcp": { "qdrant_memory": { "type": "local", "command": ["uvx", "--from", "qdrant-memory-mcp==0.3.1", "qdrant-memory-mcp", "--home", "/path/to/profile", "mcp"], "enabled": true } } }

Pi needs a separately installed MCP client/extension. In that extension, use command: uvx and the same argument list. No built-in Pi MCP support or universal extension config format is claimed. These are configuration examples. This protocol acceptance tests use the official Python MCP SDK. Host-specific evidence is recorded separately in the architecture and support boundaries and the Runtime's client evidence and limits.

Tools and boundaries

Scope is fixed by the operator's profile (user_id, agent_id). Model inputs and MCP client identity cannot override it. Knowing an ID does not grant access. Migration, export, reset and bulk deletion remain trusted operator CLI commands. See standalone maintenance. No resources or prompts are advertised. This is a local subprocess, with no public HTTP/SSE listener or remote authentication service. Stdout is reserved for protocol messages, with sanitized diagnostics on stderr.

Data flows from client → Runtime → embedding provider → Qdrant and the local SQLite ledger. With local Ollama and embedded Qdrant, memory processing stays local. Cloud embedding receives memory/query text, and cloud Qdrant receives vectors and stored payloads. The client/model may also receive recalled text. The operator must authorize retention and access and protect the profile files.

One writer lease coordinates a profile/destination. Do not overlap MCP, CLI or Hermes processes writing to it. Separate profiles or machines do not provide a distributed lock. This package does not support uncoordinated multiprocess writes, auto-extraction, an HTTP transport, or implicit credential discovery from Hermes. It preserves the Runtime's ledger, IDs and recovery semantics.

Concepts and evidence

• Sharing vs Portability vs Backup: Runtime portability guide
• Cross-harness architecture: docs/cross-harness-architecture.md
• Shared Owner status: docs/universal-owner-status.md
• Validation limits: docs/dsh-validation.md

Validation reports apply to their recorded versions, workloads and host environments. Short protocol or cloud runs do not establish 24-hour acceptance for a new release.

Verification and development

CI builds wheel/sdist, checks metadata and ownership marker. It runs installed protocol tests with temporary profiles, embedded Qdrant and a deterministic loopback embedding fixture on Linux/macOS/Windows, Python 3.11 and 3.14. No test connects to personal Qdrant, Hermes profiles or memory data. These tests cover initialize, discovery, four tools, persistence, recovery, scope isolation, error redaction and protocol stdout. Provider/cloud compatibility evidence belongs to the Runtime.

CI prefers the immutable Runtime PyPI release. Before a Runtime release reaches PyPI, CI builds a wheel from the immutable revision recorded in runtime-source.json. This is only a preparation fixture: published dependency metadata always uses the exact PyPI version. Release workflow acceptance requires installation from PyPI and never uses that fallback.

See CI and dependency maintenance, publishing, architecture, and the optional Skill.

Dependency boundary

The runtime dependency is only hermes-plugin-qdrant-memory[mcp]. The sdist includes the Python acceptance tests, apart from harnesses th