CLI
bash
npm i -g @vefimov/crew # or put this before each command: npx @vefimov/crewServer
| Command | What it does |
|---|---|
crew or crew start | Starts the server and web UI on 127.0.0.1:3000. |
crew start -p 8080 --host 0.0.0.0 | Uses another port or interface. |
crew start --no-telegram | Doesn't start the bot. |
crew dev | Starts Crew and opens the browser. |
crew doctor | Checks Node, the database, the provider and the models. |
Agents
| Command | What it does |
|---|---|
crew agent create <name> [-t template] [-r role] [--persona] [--prompt] [-m model] [-a emoji] [-f folder...] [--rw] | Creates an agent. |
crew agent ls | Lists agents with their memory counts. |
crew agent rm <name> | Deletes an agent and all of its memory. |
crew agent grant <name> <folder> [--rw] | Shares a folder. |
crew agent export <name> [-o file] | Exports to JSON. |
crew agent import <file> | Imports from JSON. |
Chat and memory
bash
crew chat Lena # interactive; /memory and /exit work inside
crew chat Lena -m "standup summary?" # one message, handy in scripts
crew memory Lena --search "deploy"
crew memory add Lena "Deploys are on Thursdays" -t decisionSettings
bash
crew config # show
crew config set model qwen2.5:7b
crew config set provider openai
crew config set openaiBaseUrl http://127.0.0.1:1234/v1