WebAgents CLI

Both SDKs install a webagents command, and it is the same command: the same subcommands, arguments, flags and messages, reading the same agent file (AGENT.md) and the same settings, keys and sign-in. A script or a habit written against one works against the other.

Start with the Quickstart.

Installation

npm install -g webagents

Everyday Commands

webagents init my-agent              # a project folder with AGENT.md
webagents                            # chat with this folder's agent
webagents -p "Summarize this"        # one prompt, answer on stdout
webagents doctor                     # what stands between this folder and a running agent
webagents serve                      # this agent over HTTP, on port 3000
webagents publish                    # send it to Robutler

Command Tree

webagents               Chat (the default command; also `chat` and `connect`)
├── serve [path]        One agent over HTTP (--port, --host)
├── daemon              Every agent in a folder, reloaded as files change (--port, --host, --watch, --no-cron)
├── init [name]         A project folder with AGENT.md (--template chatbot|tool-agent)
├── publish [path]      Create the agent on Robutler, or update the linked one (--yes, --dry-run)
├── login, logout       Your Robutler account (--url, --token)
├── whoami              Who you are signed in as
├── link [name]         Link this folder to one of your agents (--show)
├── unlink              Forget that link
├── doctor              Check this setup and say what to fix
├── models              Model providers, and which have a key here
├── skills              list, add, remove: the skills an agent file names
├── templates list      What `init --template` can make
├── config              get, set, unset, validate, path
└── secrets             list, set, unset, get: keys kept on this machine

Global flags go before the command: --json (one JSON document on standard output), --profile <name> (separate settings, keys and sign-in) and --token <token> (a platform token for this run only). -V prints the version and -h the help, for any command.

Chat

webagents opens a chat with the agent in the current folder, or the built-in assistant where there is none: it works with the files in the folder and calls web APIs. The chat has the same commands, keys and conversation files in both SDKs. See Chat.

Where the SDKs Differ

The command is the same; a few things underneath are not.

Sandbox. The Python SDK confines an agent's shell commands with the operating system (Seatbelt on macOS, bubblewrap on Linux) when the agent file declares sandbox:. The TypeScript SDK has no sandbox yet, and webagents doctor says so. See Sandbox.

Shared context. The Python loader merges WEBAGENTS.md context files into the agents below them. The TypeScript loader reads the agent file alone.

Skills. The two SDKs ship different skill sets; webagents skills list names what each can load.

Configuration

Settings are read in this order, first match wins:

command-line flags

environment variables, for the settings that have one (ROBUTLER_API_URL, WEBAGENTS_PROFILE, WEBAGENTS_TOKEN)

./.webagents/config.json in the current folder

~/.webagents/config.json

built-in defaults

Agent behavior itself lives in the AGENT.md front matter. See Configuration for the keys, and Publish for how the platform address is chosen.