English | 简体中文
Documentation · Quick start · Live demo · Contact
⭐ Help more people find LinkSense and grow the community. Star this repo!
LinkSense gives organizations one governed place to run agent tasks, share reusable Knowledge, Skills, Plugins, and Applications, and connect business systems without handing members the underlying credentials. Deploy it on your servers or private cloud, choose hosted or compatible local models, and keep task data and files on your own infrastructure.
LinkSense supports Linux and macOS hosts with Docker. A Core evaluation needs at least 4 vCPU, 8 GiB of memory available to Docker, and 60 GiB of free SSD. Full adds built-in document parsing, indexing, and retrieval and needs more capacity.
The fastest installation path is to give a coding agent this prompt:
Help me install and start LinkSense Core by following https://raw.githubusercontent.com/LingX-AI/linksense/main/Install.md
For a manual installation, choose Core or Full. The site covers host checks, Docker requirements, installation commands, and first-time setup.
- 🤖 Agent execution and deliverables: research sources, operate tools, run code, analyze data, and produce documents, spreadsheets, presentations, and other files in a user Worker.
- 🧩 Shared organizational capabilities: publish governed Knowledge, Skills, Plugins, and Applications for specific users and groups.
- 🔗 Business system connections: connect documents, email, ERP, CRM, and other systems through connectors or MCP within configured permissions.
- 🛡️ Access and governance: manage members, model credentials, quotas, usage, and audit records from one administration surface.
- 🏠 Self-hosting and model choice: use your own infrastructure, external model APIs, or compatible self-hosted endpoints with per-user runtime resource limits.
- ⚡ Flexible entry points: start work from the web workspace, messaging channels, or scheduled tasks.
| I want to… | Read |
|---|---|
| Compare deployment profiles | Core and Full |
| Operate, repair, or upgrade a deployment | Operations |
| Run tasks and use shared resources | User guide |
| Manage members, models, permissions, and usage | Administrator guide |
| Build applications or connect external systems | Developer guide |
| Work from source or understand the runtime | Source development ·Architecture |
For private deployment or business system integration, email hello@linksense.org with your environment, expected usage, and the systems you want to connect.
Read the contribution guide and repository instructions before opening a pull request. Code contributions require a Contributor License Agreement.
Questions and feature ideas belong in GitHub Discussions. Organizations can opt in to the adopters list.
Choose the channel that best fits your question:
- GitHub Discussions : Get help, share feedback, and propose ideas.
- GitHub Issues : Report reproducible bugs and track engineering work. Read thecontribution guide before opening one.
- Discord : Chat in real time, share your apps, and connect with other LinkSense users.
- X : Follow LinkSense for release news and project updates.
- WeChat group : Scan the QR code to join the LinkSense user group. It is valid for 7 days; if it has expired, emailhello@linksense.org for the latest invite.
The isolation boundary is the user Worker; tasks owned by the same user can access their shared files. Normal execution allows outbound internet access without per-command approval, and configured external services receive the data sent to them. Read the security policy before exposing a deployment. Report vulnerabilities privately to security@linksense.org.
LinkSense-owned code is licensed under CPAL-1.0. See the internal deployment exception, attribution requirements, and trademark policy. Licensing questions: licensing@linksense.org.
The upstream Codex runtime is licensed under Apache-2.0. LinkSense is an independent project; see the third-party notices.
If you find LinkSense useful or interesting, a GitHub Star ⭐️ would be greatly appreciated.
Thanks to the Linux.do community for its support.
Thanks to all the developers who contribute to LinkSense.
| @Metrolive | @guygubaby | @Bo-infocare |