What Is Ultra Code Mode in Claude Code? The "ultracode" Setting and Dynamic Workflows, Explained

Search interest in "Ultra Code Mode" has spiked since the word ultracode started showing up — highlighted in violet — inside Claude Code prompts. A quick clarification before the how-to: Anthropic's own docs and changelog never use the spelled phrase "Ultra Code Mode." The actual feature is a Claude Code setting named ultracode, and per the Claude Code workflows documentation it combines xhigh reasoning effort with automatic workflow orchestration. With it on, Claude plans a workflow for each substantive task on its own instead of waiting for you to ask. This piece is the focused explainer of how that works — not a rehash of the May 28, 2026 launch of Opus 4.8 and Dynamic Workflows, which BitsMinds already covered.

There are two ways to switch it on. To turn the whole session over to Claude's judgment, set /effort ultracode; from then on Claude plans a workflow for every substantive task in the session. (Note one wrinkle in the sourcing: the Claude Code changelog documents /effort ultracode only inside a v2.1.160 bug-fix entry — "Fixed /effort ultracode incorrectly blaming the dynamic workflows setting when the model cannot run xhigh" — rather than as the headline way to enable it; the workflows doc page is where it is described as the session-level switch.) To run a single task as a workflow without changing the session's effort level, just include the keyword ultracode in your prompt. Asking in plain words — "use a workflow" or "run a workflow" — counts as the same opt-in. According to the changelog, that trigger keyword was renamed from workflow to ultracode in v2.1.160 (June 2, 2026); the workflows doc phrases the cutover as "before v2.1.160 the literal trigger keyword was workflow," and natural-language requests work in both versions. If the violet highlight gets in your way, press Alt+W on Windows or Linux (Option+W on macOS), or backspace right after the keyword; you can disable the trigger entirely from /config.

Under the hood, this is Dynamic Workflows. As the Claude.com announcement puts it, Claude "dynamically writes orchestration scripts that run tens to hundreds of parallel subagents in a single session" — the workflows doc specifies the script is JavaScript, and a runtime executes it in the background while your session stays responsive. Claude breaks the prompt into subtasks, fans the work out across subagents, and the verification is adversarial: agents attack the problem from independent angles while others try to refute what they found, iterating until results converge. Because the loop, the branching, and the intermediate results all live in script variables rather than the conversation, Claude's context holds only the final answer — which is what lets the plan "stay on track no matter how big the task gets." The runtime allows up to 16 concurrent agents (fewer on machines with limited CPU cores) and 1,000 agents total per run to prevent runaway loops; every run writes its script to a file under ~/.claude/projects/ and is resumable within the same session. A few hard limits worth internalizing:

• No mid-run user input — only agent permission prompts can pause a run.
• Spawned subagents always run in acceptEdits mode and inherit your tool allowlist regardless of your session's mode, so file edits are auto-approved.
• Resume is session-bound — exit Claude Code mid-run and the next session restarts the workflow fresh.
• Cost and time scale up — Anthropic warns Dynamic Workflows can consume substantially more tokens than a typical session, and each session request takes longer.

On availability, the sources are mostly aligned but not identical, so it is worth stating conservatively. Dynamic Workflows are a research preview requiring Claude Code v2.1.154 or later, working across the CLI, Desktop app, IDE extensions, non-interactive claude -p, and the Agent SDK, plus the Anthropic API, Amazon Bedrock, Google Cloud Vertex AI, and Microsoft Foundry. The launch blog and InfoQ both name Max, Team, and (admin-enabled) Enterprise plans; the workflows doc additionally lists Pro, noting "on Pro, turn them on from the Dynamic workflows row in /config." Treat Pro inclusion as the doc's claim that the launch announcement does not echo. Either way, ultracode only appears on models that support xhigh — on others the /effort menu does not offer it.

So when should you reach for it? The good fits are exactly the jobs too large for a single agent: codebase-wide migrations, security audits, big refactors, performance reviews, investigating widespread bugs, and deep research — work the blog says can stretch into hours and days. Claude Code even ships /deep-research as a bundled workflow that fans web searches across angles, cross-checks sources, votes on each claim, and filters out the ones that do not survive. When not to use it is just as important: for routine edits, the workflows doc is explicit — "drop back with /effort high when you return to routine work." InfoQ adds the sensible on-ramp of starting with smaller, well-scoped tasks before pointing it at a large project. And remember the relationship to the May 28 launch: Dynamic Workflows shipped that day in v2.1.154 alongside Opus 4.8; ultracode is the convenience layer added on top — the rename to the ultracode keyword and the /effort ultracode wiring landed later, in the June 2 v2.1.160 release.