coneshare

mcp
Security Audit
Fail
Health Pass
  • License — License: MIT
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Community trust — 49 GitHub stars
Code Fail
  • Hardcoded secret — Potential hardcoded credential in .github/workflows/build-ci.yml
Permissions Pass
  • Permissions — No dangerous permissions requested

No AI report is available for this listing yet.

SUMMARY

Self-Hosted Virtual Data Room (VDR) & Document Tracking. Open-Source DocSend alternative with Built-in MCP Server.

README.md

English | 中文

Coneshare logo

Build CI
Ask DeepWiki
MIT License
Documentation

Coneshare

Coneshare is an open-source, self-hosted alternative to DocSend that works on top of your existing storage (Nextcloud, Google Drive, or Dropbox). Files stay on your own servers instead of moving to third-party cloud providers.

Use Coneshare to share documents and videos securely: apply dynamic watermarks, track viewer engagement page by page in real time, and trigger automations from viewer activity.

Quick start · Docs · Live demo · Roadmap · Forum

If you find this project useful, please consider starring the repository.

▶ Click here to watch the 2-minute product walkthrough video


How it works

Coneshare runs directly on top of your storage:

  • Files remain in your storage provider (Nextcloud, Google Drive, or Dropbox).
  • Coneshare adds:
    • Virtual datarooms and media previews
    • Access controls and link permissions
    • Document engagement tracking
    • Webhooks and workflow automation

Keep your existing storage workflows while gaining access controls, datarooms, and engagement analytics.


Features

Virtual datarooms and media previews

  • Organize documents and rich media into structured rooms.
  • PDF rendering and secure video streaming.
  • Inline viewers to navigate between related items.

Access control

  • Password protection, link expiration dates, and email verification.
  • Download restrictions and dynamic watermarks.

Analytics and engagement tracking

  • Track views, revisits, and downloads in real time.
  • View page-by-page reading duration and video playback metrics.

Automation and webhooks

  • Slack and webhook notifications for link events.
  • Trigger actions based on viewer activity.

Infrastructure, privacy, and internationalization

  • Self-hosted on your own servers or private cloud.
  • Available in English, Simplified Chinese, Russian, German, French, and Dutch (en, zh-hans, ru, de, fr, nl).

AI agents and remote MCP

You can connect web AI assistants (ChatGPT, Gemini, Claude.ai, and Grok) and CLI or desktop coding agents (Claude Code, Claude Desktop, Cursor, VS Code, Codex, and Antigravity agy) to your Coneshare workspace over the remote Model Context Protocol (MCP) endpoint (/mcp/sse). The server provides MCP tools for saving AI conversations, organizing datarooms, creating protected or watermarked share links, and checking viewer engagement.

  • Web AI clients sign in with OAuth 2.1 and PKCE (read or write scopes). Destructive deletes, admin routes, and viewer PII remain blocked.
  • CLI and desktop agents authenticate with scoped API keys (read_only, read_write, or full_access).

See mcp-server/README.md for setup steps, the walkthrough video, and the tool list, or check Coneshare AI & MCP and MCP Docs.


Integrations

Coneshare connects with:

  • Nextcloud (self-hosted)
  • Google Drive
  • Dropbox

More storage connectors are in development.


Deployment (self-hosted)

For production deployments, use the coneshare-compose repository. It provides Docker Compose configurations with automated Let's Encrypt SSL, reverse proxy setups, and preconfigured containers.


Local development

To run the codebase locally for development and contributions:

Source build

git clone [email protected]:coneshare/coneshare.git
cd coneshare
cp .env.template .env
make build
make up
make migrate

First-run verification checklist

After make up and make migrate, verify the local services before configuring storage integrations:

  • Frontend is reachable at http://localhost:5173
  • API responds at http://localhost:8000/api/v1/
  • Core services are up (backend, frontend, core, redis, celery)
  • Local files persist under the project data and storage volumes defined in docker-compose.yml
  • Smoke test:
    • Upload one document
    • Create a share link
    • Open the link in a private browser window to confirm view access works

Troubleshooting first install

  • .env and SITE_DOMAIN mismatch:
    • Confirm .env exists (copied from .env.template) and SITE_DOMAIN matches how you access the app locally.
  • Backend cannot reach core service:
    • Check service names and ports in docker-compose.yml, then inspect backend and core logs for connection errors.
  • Redis or Celery issues (background jobs not running):
    • Confirm both redis and celery containers are running, then check Celery worker logs.
  • Storage permissions:
    • Verify mounted storage directories exist and are writable by container processes.
  • Where to look first:
    • Run make logs from the repository root, checking backend, core, and celery error output.

Architecture

Coneshare architecture diagram

Coneshare consists of four primary services:

  • backend/: Django and DRF API, Celery, and Redis for background tasks
  • core/: Go service for file delivery and media streaming
  • frontend/: React and Vite web application
  • mcp-server/: FastMCP remote server for AI agent and assistant integrations

Technical reference:


Contributing

Contributions are welcome. Please read CONTRIBUTING.md for our code of conduct and pull request process.

Contributors

Justin Zheng
Justin Zheng

💻 🚇 📖 🎨
James Ramsay
James Ramsay

💻 ⚠️ 🐛
Matthias Friedewald
Matthias Friedewald

💻 ⚠️ 🐛
KeizerDev
KeizerDev

💻 🌍 ⚠️
Tom Elliot
Tom Elliot

💻 ⚠️ 🐛
Agatino Platania
Agatino Platania

💻 ⚠️ 🐛
Industry
Industry

💻 ⚠️ 🐛

Community


License

MIT License. See LICENSE.


Security

For security issues, contact [email protected].

Reviews (0)

No results found