cc-fleet

Run any model with an Anthropic- or OpenAI-compatible API — DeepSeek, GLM, Kimi, Qwen, MiniMax, even your Codex subscription — as scripted multi-agent workflows, real Claude Code agent-team teammates, or one-shot subagents, driven just like native ones. Your main session's own auth (OAuth subscription or API key) is untouched; each worker bills its own provider — an API key, or your Codex subscription.

No Claude subscription? ccf run <provider> starts an interactive session on that provider — the same claude you know, just on the provider's model.

Install

npm install -g @ethanhq/cc-fleet
# or run without installing:
npx @ethanhq/cc-fleet --help

postinstall downloads the prebuilt release archive for your platform (linux/darwin/windows × x64/arm64) from the matching GitHub Release, verifies its sha256 against checksums.txt, then unpacks the binary — giving you cc-fleet and the ccf alias.

The skill

The npm package installs only the CLI binary. To teach Claude Code when to reach for the fleet — scheduling Workflows, Agent Teams, and Subagents on the right lane — install the cc-fleet plugin. Either way:

• Inside Claude Code (recommended) — launch claude and run the two slash commands (or /plugin for the interactive panel):

  /plugin marketplace add ethanhq/cc-fleet
  /plugin install cc-fleet@ethanhq

• From the shell:

  claude plugin marketplace add ethanhq/cc-fleet
  claude plugin install cc-fleet@ethanhq

First run

cc-fleet             # open the TUI and register a provider (config created on first save)

Common commands

cc-fleet and ccf are the same binary — use whichever you prefer.

ccf doctor               # health-check (core vs optional)
ccf list                 # list configured providers
ccf run <provider>       # interactive claude session on that provider
ccf update               # update the binary + plugin (--check to only check)
ccf uninstall            # reset cc-fleet state (--all to fully remove)

Full documentation: https://github.com/ethanhq/cc-fleet