ethanhq/cc-fleet

Subagent fans out one-shot or parallel cc-fleet provider workers that return a result. Use it for delegated research, bulk per-file work, or a single background model call, not for multi-phase workflows or local editing.

Subagent runs one-shot or parallel headless provider LLM jobs and returns the result on stdout. Use it for fan-out tasks, bulk per-file work, or calling a specific provider model without a long-lived conversation.

Team spawns long-lived provider LLM teammates in tmux panes and lets you message them across turns. Use it for sustained parallel work, collaborative debugging, or watchable multi-step tasks.

Workflow orchestrates multi-phase cc-fleet runs over provider subagents from a JavaScript script. Use it for fan-out, barriers, per-item pipelines, resumable runs, or other dependent workflows that need to run off the main context.

Workflow orchestrates multi-phase, dependent, or resumable runs over many provider subagents from a JS script off the main context. Use it for fan-out/barrier/synthesis, per-item pipelines, or runs that must resume after interruption.