Grok Build now supports workflows, allowing users to describe complex tasks in plain language that are automatically planned and executed across hundreds of parallel agents. These orchestration scripts fan out work, verify results, and report back in a single background run while keeping the user's session free.
- Runs support up to 128 agents by default, with an option for 1,024 agents for large jobs.
- Progress is saved during execution, enabling pausing and resuming without redoing finished work.
- Users can monitor runs live via /workflows, viewing phase-by-phase progress and per-agent token counts.
- Saved workflows in .grok/workflows/ or ~/.grok/workflows/ become reusable slash commands with arguments.
- The /deep-research feature utilizes this infrastructure to fan out research questions to parallel investigators.
This capability is designed for multi-faceted tasks that exceed the context of a single conversation, such as auditing codebases or triaging issues, by ensuring independent verification of findings before final reporting.