Divlens
Local MCP server giving AI clients live CPU, memory, disk and hardware diagnostics
- Pricing
- Open source
- Platforms
- macOS, Windows, Linux
- Version
- 0.1.0
- Listing updated
- today
Similar MCP servers
About Divlens
Divlens is an open-source MCP server for developer tools. It is available on macOS, Windows, Linux and CLI.
The README describes DivLens as a Rust-based Model Context Protocol server that runs on your own machine and lets an AI client query live system state. Instead of guessing why a laptop is slow, an assistant can request CPU and RAM figures, the process list, storage usage, SMART status, network throughput and recent OS or kernel errors through named tools.
The server speaks newline-delimited JSON-RPC over stdio, so it attaches to clients such as Claude Desktop, Cursor and Windsurf through a short config entry. The README states that DivLens does not run any language model itself; it only collects and returns raw data, and that nothing leaves the machine. An optional vector-memory build adds semantic recall of past diagnoses using a local embedding model.
It is aimed at developers and power users who want an assistant to troubleshoot their own computer. The README notes that the installer script covers macOS, Linux and Windows, and that a source build requires Rust 1.82 or later.
Tools this server provides (17)
Taken from the project documentation; the running server may differ.
- get_live_metrics
- CPU %, RAM, swap, blocked processes, uptime
- get_process_list
- Top processes by CPU / RAM with PID
- get_storage_health
- Free/used/total per mount point
- scan_storage_inventory
- Full file-type inventory with sizes
- get_file_type_summary
- File counts and sizes by extension
- get_specific_file_type
- All files matching a specific extension
- get_advanced_storage_stats
- Top 50 largest files + stale data analysis
- get_storage_diagnostics
- IOPS, read/write latency, SMART status
- get_hardware_diagnostics
- CPU/GPU specs, battery %, temps, SMART
- get_network_diagnostics
- Throughput, active connections, signal
- get_network_config
- IP, DNS, interface config per adapter
- get_system_dna
- OS, hostname, uptime, machine fingerprint
- get_dev_stack
- Node, Python, Rust, Go, Java runtimes + packages
- get_drivers
- Kernel modules and device drivers
- scan_directory
- Recursive directory listing with sizes
- recall_memory
- Semantic search over past AI diagnoses
- get_system_logs
- Recent OS/kernel errors clustered by pattern
What it does
- Live system metrics
- Returns CPU usage, memory, swap, uptime and the top processes by CPU or RAM.
- Storage inspection
- Reports free and used space per mount point, largest files and file-type breakdowns.
- Hardware and SMART checks
- Surfaces CPU and GPU specs, battery level, temperatures and disk health data.
- Network diagnostics
- Shows throughput, active connections and per-adapter IP and DNS configuration.
- Developer runtime inventory
- Lists installed Node, Python, Rust, Go and Java runtimes and their packages.
- Local-only operation
- Runs as a local stdio process with no cloud calls or API keys, according to the README.
Good for
- Asking an assistant why a Mac or PC is running slowly
- Checking whether an SSD or disk shows signs of failure
- Finding which folders and files are using the most disk space
- Reviewing recent OS or kernel errors without digging through logs
- Confirming installed developer runtimes and versions on a workstation
Pricing
Open source
Apache License 2.0; no paid plan mentioned.
Divlens FAQ
Does DivLens send my system data to the cloud?
The README says all data stays on the machine and nothing is sent anywhere, and it describes the server as running with no cloud component.
Which AI clients does it connect to?
The README gives configuration steps for Claude Desktop, Cursor and Windsurf, and says any MCP-compatible client can use it.
Does it need an API key or Rust installed?
No API key is required. The installer downloads a prebuilt binary, so Rust is only needed for building from source.
Is semantic memory enabled by default?
No. The README says recall_memory returns an empty list unless the vector-memory feature is compiled in.
Is Divlens free?
Yes. Divlens is open source, so you can use and self-host it for free.
What are the best alternatives to Divlens?
The closest alternatives to Divlens are Lens MCP, Linux MCP, codeDB Lens, ILens and Codelens MCP.
Where this comes from
- Description: written from the project's README and its MCP Registry entry.