skills/castari-deploy/SKILL.md
This skill should be used when the user asks to "deploy to castari", "deploy my agent", "deploy agent to production", "castari deploy", "push agent to castari", "ship my agent", "set up castari", or needs help with Castari CLI installation, authentication, agent scaffolding, or deployment to cloud sandboxes.
npx skillsauth add castari/cli castari-deployInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Deploy a Claude AI agent to production on Castari. This skill walks through the full flow: CLI installation, authentication, project scaffolding, and deployment.
Follow these steps in order. Skip any step whose prerequisite is already satisfied.
Run:
cast --version
If the command fails or is not found, install the CLI:
npm install -g @castari/cli
Then verify installation succeeded by running cast --version again.
Run:
cast whoami
If the user is not logged in (command fails or shows "not authenticated"), run:
cast login
This opens a browser for Clerk OAuth. Wait for the user to complete login, then verify with cast whoami again.
Look for castari.json in the current working directory:
ls castari.json
If castari.json exists: Read it and confirm the configuration with the user. Proceed to Step 4.
If castari.json does not exist: Help scaffold one:
Detect the likely entrypoint by checking for these files in order:
src/index.tssrc/main.tsindex.tssrc/index.jssrc/main.jsindex.jsAsk the user for the agent name using AskUserQuestion. Suggest the current directory name as the default.
Generate castari.json with this structure:
{
"name": "<agent-name>",
"version": "0.1.0",
"entrypoint": "<detected-entrypoint>",
"runtime": "node"
}
Write the file and show the user what was created.
Run the deployment:
cast deploy
This packages the project, uploads it, and deploys to an isolated cloud sandbox. The command may take up to a minute. Show the user the output including status and sandbox ID.
After successful deployment, show the user a summary:
Ask the user if they want to test the deployed agent. If yes, ask for a test prompt and run:
cast invoke <agent-slug> "<test-prompt>"
cast login again or checking network connectivity.castari.json configuration or running cast agents list to verify the agent exists.castari.json file is the project manifest — it must be in the repo root.src/index.ts for TypeScript projects, index.js for JavaScript.tools
Use when work should span one or more detached tasks but still behave like one job with a single owner context. TaskFlow is the durable flow substrate under authoring layers like Lobster, ACPX, plugins, or plain code. Keep conditional logic in the caller; use TaskFlow for flow identity, child-task linkage, waiting state, revision-checked mutations, and user-facing emergence.
tools
# Lobster Lobster executes multi-step workflows with approval checkpoints. Use it when: - User wants a repeatable automation (triage, monitor, sync) - Actions need human approval before executing (send, post, delete) - Multiple tool calls should run as one deterministic operation ## When to use Lobster | User intent | Use Lobster? | | ------------------------------------------------------ | --------------------------
tools
# Lobster Lobster executes multi-step workflows with approval checkpoints. Use it when: - User wants a repeatable automation (triage, monitor, sync) - Actions need human approval before executing (send, post, delete) - Multiple tool calls should run as one deterministic operation ## When to use Lobster | User intent | Use Lobster? | | ------------------------------------------------------ | --------------------------
tools
A CLI tool for making authenticated requests to the X (Twitter) API. Use this skill when you need to post tweets, reply, quote, search, read posts, manage followers, send DMs, upload media, or interact with any X API v2 endpoint.