Skip to main content

Persistent knowledge graph memory for AI agents — DejaView MCP server

Project description

dejaview-mcp

Persistent knowledge graph memory for AI agents — an MCP server backed by DejaView.

Connect Claude Desktop, Cursor, Windsurf, or any MCP-compatible host to a live knowledge graph. Your AI remembers people, projects, decisions, and relationships across every session.

Install

pip install dejaview-mcp

Get a free API key at dejaview.io.

Quick start

DEJAVIEW_API_KEY=dv_your_key_here dejaview-mcp

Configure Claude Desktop

Edit your Claude Desktop config file:

  • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
  • Linux: ~/.config/claude/claude_desktop_config.json
  • Windows: %APPDATA%\Claude\claude_desktop_config.json
{
  "mcpServers": {
    "dejaview": {
      "command": "dejaview-mcp",
      "env": {
        "DEJAVIEW_API_KEY": "dv_your_key_here"
      }
    }
  }
}

Restart Claude Desktop. DejaView will appear in the tools panel.

Configure Cursor / Windsurf

{
  "mcpServers": {
    "dejaview": {
      "command": "dejaview-mcp",
      "env": {
        "DEJAVIEW_API_KEY": "dv_your_key_here"
      }
    }
  }
}

What it does

DejaView gives your AI a persistent knowledge graph it can read from and write to across sessions. Unlike flat context windows that reset every chat, the graph grows over time.

Tool What it does
agent_context() Load a full memory summary at session start
remember(subject, predicate, object) Store a fact
remember_many(facts) Store multiple facts at once
recall(entity) Get everything known about an entity
search(query) Find entities by name
ask(question) Natural language Q&A over the graph
timeline() See recently stored facts
graph_stats() Entity/relationship counts
share(entity) Generate a public shareable link
forget(subject, predicate, object) Remove a specific fact
forget_entity(name) Remove an entity and all its connections

Example

Once connected, just chat naturally:

"Remember that Alice is the lead on Project Atlas and prefers async communication."

The agent calls remember() automatically. Next session:

"What do I know about Alice?"

The agent calls recall("Alice") and tells you everything — including things you told it months ago.

Self-hosting

Want to run your own DejaView instance? The API is open source: github.com/JakeC77/DejaView

Set DEJAVIEW_ENDPOINT to point at your instance:

DEJAVIEW_API_KEY=dv_... DEJAVIEW_ENDPOINT=https://your-instance.com dejaview-mcp

Links

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distributions

No source distribution files available for this release.See tutorial on generating distribution archives.

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

dejaview_mcp-0.1.0-py3-none-any.whl (7.8 kB view details)

Uploaded Python 3

File details

Details for the file dejaview_mcp-0.1.0-py3-none-any.whl.

File metadata

  • Download URL: dejaview_mcp-0.1.0-py3-none-any.whl
  • Upload date:
  • Size: 7.8 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.3

File hashes

Hashes for dejaview_mcp-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 03378211ca618bbd60c723c4e967d3a9b7638c8aa4db759f099e9b8d2661831d
MD5 bf4c98eb040707a60331dd22ddf88881
BLAKE2b-256 74f09926ad2f355fc9b64366eac648b3f69ef77b0c3a7ad002fe9b1c09d11067

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page