Docent- An Open Source Private AI assistant in your terminal

Smartloop released Docent, an open-source private AI assistant that runs in the terminal and lets users chat with PDFs and Office files, search the web, and connect MCP servers, with the CLI automatically downloading the SLP framework 1.2.8 and its models when no local agent is running. The tool installs as both `docent` and `smartloop`, ships prebuilt binaries for Linux (x86_64, aarch64, statically linked against musl), macOS (Apple Silicon and Intel) and Windows (x86_64), and on first use downloads the bge-m3-Q4_K_M.gguf embedding model (~417 MB) plus a default base model into ~/.smartloop/1.2.8/.

Your private AI assistant. Chat with your PDFs and Office files, search the web, and connect MCP servers, all from the terminal. If no local agent is running, the CLI downloads the framework and models it needs and starts one itself. It installs as both docent and smartloop ; the two are the same command. More at: docs.smartloop.ai https://smartloop.ai/docs/intro/ macOS and Linux curl -fsSL https://smartloop.ai/install | sh Windows PowerShell : irm https://smartloop.ai/install.ps1 | iex The binary goes to $CARGO HOME/bin ~/.cargo/bin when a Rust toolchain already owns that directory, since it is on your PATH anyway; otherwise to ~/.local/bin . If the chosen directory is not on your PATH , the installer appends an export line to the startup file for your shell — ~/.bashrc , ~/.zshrc , or fish add path in config.fish — so a new terminal picks it up. Re-running the installer will not add that line twice. On Windows the binary goes to %USERPROFILE%\.smartloop\bin and the installer sets your user PATH through the registry. Restart the shell to pick it up. Set SMARTLOOP CLI INSTALL DIR to install elsewhere, or SMARTLOOP CLI VERSION to pin a specific release. Prebuilt binaries are published for Linux x86 64, aarch64 — statically linked against musl , macOS Apple Silicon and Intel and Windows x86 64 . Requires Rust 2024 edition : cargo install --path . Any command that talks to the agent starts it when none is running. It listens on a free port it picks itself and writes it to ~/.smartloop/server.port , where the CLI reads it. On first use that means: 1. Download the agent SLP framework 1.2.8 from https://dl.smartloop.ai/slp/1.2.8/ into ~/.smartloop/1.2.8/ . Studio desktop uses the same folder and marker files, so the two share one install. 2. Start slp agent start in the background with SLP HOME=~/.smartloop , logging to ~/.smartloop/server.log . The agent keeps running after the CLI exits. 3. Download the embedding model bge-m3-Q4 K M.gguf , ~417 MB into the workspace's models/embeddings/ folder for document search. 4. Run the agent's bootstrap, which downloads the default base model, creates the default project and loads the model. Each step is skipped when its files are already there. Progress shows as a checklist on stderr under the Docent banner that redraws in place, with the active download's bar in Smartloop pink: ✓ Agent 1.2.8 667 MB ✓ Start agent port 50578 ✓ Embeddings bge-m3 417 MB • Base model sl-mini ██████████████▋░░░░░░░░░░░░░░░ 49% 377 MB/769 MB Default project Load model Skills and connections A step without a bar shows its elapsed time once it runs past a couple of seconds, and it ends with ✓ Setup complete in 1min 12s . When stderr isn't a terminal, each step prints one line as it finishes instead. docent model enable shows the same checklist for its download. To do this up front, or to stop the agent: docent agent start docent agent stop docent login Opens app.smartloop.ai in the browser to sign in, the same way the desktop app does: once you're in, the page hands the session to the local agent, which stores it and uses it for every platform call. The command waits up to five minutes and prints the account it signed in as. If the browser doesn't open, visit the URL it prints. To sign in with a token instead, run docent login --token and paste it at the prompt input is hidden , or use --token <token or pipe it on stdin in scripts. docent logout clears the stored credentials. List projects: docent project list Output is rendered as a table showing each project's ID, name, and whether it is a system project. Create a project from a blank template: docent project create --name my-project docent project create --name my-project --description "Research notes" A blank project starts with no skills; the service seeds it with the workspace defaults. Everything the project stores — skills, documents, its index — lives under the service's own project directory, so there is no working directory to choose. Import a project from an archive produced by an earlier export: python docent project create --import my-project.zip docent project create --import my-project.zip --name restored-project --name is optional here — pass it to rename the imported project. The import gets a fresh project ID, and MCP OAuth credentials are stripped from the archive on the way in. Delete a project: docent project delete --id <project-id Check which endpoint the CLI uses, whether the local agent is running there, which model it has loaded, and the per-project agents it has started: docent agent status Endpoint: http://localhost:50578 Status: healthy Model: sl-mini Q4 K M, 32768 ctx, 769 MB Process: pid 8738, 250 MB +--------------+------+-------+-------+------+--------+ | Project | PID | Port | Alive | Idle | Memory | +--------------+------+-------+-------+------+--------+ | general chat | 2668 | 62110 | true | 94s | 122 MB | +--------------+------+-------+-------+------+--------+ Endpoint is the URL the CLI talks to SMARTLOOP API URL , or the default . The command exits with status 1 when the agent can't be reached there. Start an interactive chat with the local agent: docent docent on its own is short for docent run . In a terminal this opens a full-screen app, laid out like Claude Code. While a reply runs, a live status line above the prompt shows the agent's current step web search, document lookup, model selection and for how long, with a line per running download. The finished reply keeps its sources and the model that answered: what is the capital of France? ■ Paris is the capital of France. References 1 https://en.wikipedia.org/wiki/Paris ⎿ sl-mini · 7 tokens · 0.4 tok/s · 16s - Reading en.wikipedia.org… 4s ╭──────────────────────────────────────────────────────────────────────╮ │ │ ╰──────────────────────────────────────────────────────────────────────╯ enter send esc interrupt ? shortcuts It chats in the server's current project, or the one given with --project . On exit it prints the session id, to resume with --session . | Key | Does | |---|---| | ? | shortcuts | | Shift+Enter , Alt+Enter , Ctrl+J | new line | | Esc | interrupt the reply | | wheel, PgUp / PgDn | scroll; drag to select and copy | | drop a file | attach an image PNG, JPEG or document PDF, Office, CSV, text | | Ctrl+O | models: enable, disable, download | | Ctrl+S | web search on or off | | Command | Does | |---|---| | /login , /logout | sign in in the browser --token to paste one | | /models | enable, disable and download models | | /mcp | MCP servers; /mcp add <url connects one | | /status | account, model, agent and versions | | /usage | web searches left, tokens and cost saved | | /upgrade | Pro plan: 1,000 web searches a month | | /clear | clear the chat and start a new session | | /help , /quit | | Pass an initial prompt to send immediately: docent run "what are some things to do in madrid spain?" Options: docent run --project <project-id chat in this project docent run --session <session-id resume an existing session; a new one is created when omitted docent run --plain line-by-line chat instead of the full-screen app When stdin or stdout isn't a terminal, or with --plain , run chats line by line instead: without --project it lists your projects and asks which one to chat in press Enter for the current one , then keeps reading prompts from stdin until EOF, /quit , /exit , /q , or exit . The response streams token by token on stdout, and the agent's progress is printed on stderr as step message lines, so it doesn't interleave with the answer. When the answer draws on documents or web pages, the sources it used follow it on stdout as a numbered list: References 1 https://releases.rs/ The orchestrator picks which model serves each turn from the models enabled for the project. List what's available and its state for the current project: docent model list Enable or disable a model: docent model enable gemma4-e2b docent model disable gemma4-e2b enable downloads the weights first when they aren't on disk yet, showing progress, and only then switches the model on. Weights are shared across projects, so each model downloads once. Models marked sign in need docent login first. disable also unloads the model from memory. All three take --project <project-id to act on a project other than the current one. The sl-mini orchestrator is always on and isn't listed. The CLI connects to the local agent on whatever port it picked ~/.smartloop/server.port . Point it elsewhere with SMARTLOOP API URL : SMARTLOOP API URL=http://localhost:9000 docent project list With SMARTLOOP API URL set, the CLI only connects: it never installs or starts an agent for that URL. | Variable | Default | Effect | |---|---|---| | SLP HOME | ~/.smartloop | Where the framework, workspace, models and logs live | | SLP PORT | a free port | Pin the managed local agent to this port | | SLP BASE URL | https://dl.smartloop.ai | Where the framework archive is downloaded from | Released under the MIT License https://github.com/smartloop-ai/docent/blob/main/LICENSE .