●WORKFLOWS — Dynamic workflows are GA in Claude Code: Claude plans the work, runs hundreds of parallel subagents in one session, and verifies outputs before reporting back●SCALE — Paired with Opus 4.8, Claude Code can carry codebase-scale migrations across hundreds of thousands of lines from kickoff to merge●ULTRACODE — The ultracode setting, via the effort menu, sets effort to xhigh and lets Claude decide automatically when to use a workflow●EFFORT — claude.ai adds an effort control beside the model selector; pick extra (xhigh in Code) or max for harder, long-running tasks●WFSIZE — A Dynamic workflow size setting in /config controls how large Claude makes workflows (small, medium, or large agent counts)●OPUS48 — Claude Opus 4.8 remains generally available, improving on 4.7 across coding, agentic skills, reasoning, and knowledge work●WORKFLOWS — Dynamic workflows are GA in Claude Code: Claude plans the work, runs hundreds of parallel subagents in one session, and verifies outputs before reporting back●SCALE — Paired with Opus 4.8, Claude Code can carry codebase-scale migrations across hundreds of thousands of lines from kickoff to merge●ULTRACODE — The ultracode setting, via the effort menu, sets effort to xhigh and lets Claude decide automatically when to use a workflow●EFFORT — claude.ai adds an effort control beside the model selector; pick extra (xhigh in Code) or max for harder, long-running tasks●WFSIZE — A Dynamic workflow size setting in /config controls how large Claude makes workflows (small, medium, or large agent counts)●OPUS48 — Claude Opus 4.8 remains generally available, improving on 4.7 across coding, agentic skills, reasoning, and knowledge work

TAG

Unattended runs

2 articles

← Back to all tags

Claude Code¹ axScreenReader¹ Logging¹ Terminal¹ MCP¹ Claude API¹ Contract Testing¹ TypeScript¹

⟐ Claude Code/2026-07-15Advanced

Keeping Unattended Logs Readable — Measuring axScreenReader's Plain-Text Rendering Before Adopting It

Piping Claude Code into a log file leaves you with escape codes and redraw debris instead of an answer. Here is how I enabled axScreenReader across its three entry points, measured the difference on one real task, and moved my unattended runs over to it.

⬡ API & SDK/2026-07-08Advanced

Contract-Test Every Tool Before You Submit or Automate an MCP Connector

A connector that works once in a chat can still break silently in an unattended job through misread response shapes or double-fired writes. Here is a small harness that machine-checks tool descriptions, response contracts, idempotency, and latency, with measured numbers.