Radar / AI coding tools / Claude Code 2.1.319
Claude Code 2.1.319: background MCP tasks and deny-instead-of-prompting
Anthropic's coding agent now runs MCP task calls in the background and reports back on completion, adds a denyInsteadOfPrompting permission mode for unattended runs, and tightens auto-mode safety checks.
Why it matters
Unattended agent runs stop dying at permission prompts: a call that would wait is denied and the turn carries on, which is what makes overnight coding jobs actually finish.
What you could build with it
A DevOps team could run nightly Claude Code jobs in CI with denyInsteadOfPrompting on, letting safe refactors land while risky calls get denied and logged, then review the denial log each morning to tune the allow-list.
Does it hold up?
Shipping to every Claude Code install now; early write-ups confirm the background-task and deny behaviours work as documented, though a backgrounded task that never calls back leaves its result silently absent.
Built with Claude Code 2.1.319
- Claude Code Now Runs MCP Tasks in the Backgrounddev.to · Explains the new backgrounding: MCP task calls move to the background immediately instead of after a two-minute wait, with /tasks to list and cancel them.
- Claude Code 2.1.319 updates MCP tasks and permissionstastedev · Short release rundown: backgrounded MCP tasks, denyInsteadOfPrompting, auto-mode server checks and permission-bypass fixes.
- Claude Code 2.1.319: check refusals before trusting a finished jobsysseder · Warns that denied actions in auto mode can end a turn silently and that non-interactive runs now exit non-zero on refusals, so scripts should check exit codes.
Learn more
First spotted on github: source.
More AI coding tools
Astryx: Meta's agent-ready design system, open sourceMeta open-sourced the React 19 + StyleX design system behind 13,000 internal apps: 150+ accessible components, themes…coding · JEV 0.76GitHub Spec Kit — spec-driven development for coding agentsGitHub's open-source toolkit that gives any coding agent a structured spec, plan, tasks, implement workflow, so specs…coding · JEV 0.75Superpowers: agentic skills framework that makes coding agents plan, test, and review before shippingobra/superpowers installs composable skills into Claude Code, Cursor, Codex and eight more harnesses, enforcing a…coding · JEV 0.74Taste Skill: agent skills that give coding agents design tasteAgent skills that stop coding agents generating generic pages, adding art direction, reference images, and a browser…coding · JEV 0.74Claude Code Game Studios: turn one Claude Code session into a 49-agent game studioOpen-source Claude Code template with 49 specialist agents, 74 skills, hooks and document templates organized as a…coding · JEV 0.74gstack: Garry Tan's exact Claude Code setup23 opinionated tools that serve as CEO, Designer, Eng Manager, Release Manager, Doc Engineer, and QA inside Claude Code.coding · JEV 0.73
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.