Help us improve
Share bugs, ideas, or general feedback.
From grimoire
Guides single-tasking to eliminate context-switching costs and recover cognitive capacity. Use when multitasking or work quality feels inconsistent.
npx claudepluginhub jeffreytse/grimoire --plugin grimoireHow this skill is triggered — by the user, by Claude, or both
Slash command
/grimoire:apply-single-taskingThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Work on one task at a time with complete attention, eliminating task-switching to recover cognitive performance lost to context fragmentation.
Schedules uninterrupted 90-120 minute single-task cognitive work sessions to maximize focus and output quality for demanding tasks like coding, writing, and analysis.
Combines Pareto prioritization (80/20), timeboxing, and deep work techniques to manage attention, eliminate context-switching, and maximize high-impact output. Useful for time management, combating procrastination, and planning schedules.
Formats all responses for readers with ADHD: leads with concrete actions, numbers multi-step work, restates state across turns, suppresses tangents, gives time estimates, and makes wins visible.
Share bugs, ideas, or general feedback.
Work on one task at a time with complete attention, eliminating task-switching to recover cognitive performance lost to context fragmentation.
Adopted by: Attention research at UC Irvine (Gloria Mark), MIT (Earl Miller), and University of Michigan (David Meyer) consistently demonstrates that multitasking is a myth — the brain does not process multiple cognitively demanding tasks in parallel. Cal Newport's Deep Work (2016) built the practitioner framework on this research. Single-tasking is explicitly recommended by Google's productivity programs and McKinsey's leadership development curriculum.
Impact:
Why best: The brain has one executive attention system that serializes tasks. What feels like multitasking is rapid task-switching, which incurs a "switch cost" — a cognitive loading period every time focus changes. Single-tasking eliminates switch costs by holding attention on one context until a natural completion point.
When you switch from Task A to Task B, part of your attention remains on Task A — this is attention residue (Sophie Leroy, University of Washington, 2009). Even if Task A is "done," unresolved questions about it leak into Task B's cognitive space.
Implication: interrupting yourself between tasks (to check Slack, glance at email) creates residue even when the check takes 30 seconds. The cost is measured in the quality of the resumed task, not the duration of the interruption.
Before starting any work:
The isolation step is not optional — the presence of unread notifications creates anticipatory distraction even when you don't act on them (Stothart et al., 2015: receiving a notification, even without checking, reduces task performance).
The optimal time to switch tasks is at a natural stopping point — a completed section, a logical pause, a finished sub-task. Switching mid-sentence, mid-function, or mid-analysis maximizes residue.
If an interruption arrives (someone stops by, a message appears):
The note removes the open loop from working memory, eliminating anticipatory residue.
Not all tasks require deep single-tasking. Routine tasks (email, Slack, admin) can be batched:
Batching clusters similar cognitive modes together, reducing the switch cost between heterogeneous tasks throughout the day.
Willpower alone cannot sustain single-tasking in a notification-rich environment. Make distractions structurally harder:
| Intervention | Mechanism |
|---|---|
| Phone in another room or bag | "Out of sight" reduces urge frequency by 2× (Ward et al., 2017) |
| Website blockers (Freedom, Cold Turkey) | Remove volitional choice from distraction |
| Status set to "Do Not Disturb" | Signals unavailability without conflict |
| Single-app fullscreen mode | Removes visual cues for other apps |
| Headphones (even without audio) | Social signal for focus; reduces interruptions |
Keeping Slack/email open "just in case": The presence of the unread badge impairs task performance even without checking. Close the app.
Calling rapid task-switching "multitasking": The brain does not multitask; it switches rapidly and pays the residue cost on every switch. Recognition that it's happening is the first step to stopping it.
Treating single-tasking as only for deep work: Shallow tasks also benefit from single-tasking. Processing email with a Zoom call in the background produces worse email and worse Zoom presence.
Believing you are the exception: High multitaskers consistently overestimate their multitasking ability. The Stanford study found that heavy multitaskers performed worse on every metric — they were not skilled at it; they were simply addicted to it.