principle-never-block-on-the-human
cursor/plugins
Proceed with reversible work without asking permission; reserve confirmation for irreversible actions only.
What is principle-never-block-on-the-human?
A decision-making principle for AI agents to avoid blocking on human confirmation for reversible tasks like code changes. Apply this when tempted to ask for approval on work that can be easily undone or reviewed afterward, keeping the agent unblocked and the human asynchronous.
- Distinguishes reversible actions (code edits, notes, task splits) that should proceed without confirmation
- Identifies irreversible actions (force-push, production deletes, external messages) that require explicit approval
- Eliminates permission-pause bottlenecks by letting humans course-correct after results are presented
- Enables asynchronous human supervision by making decisions and explaining reasoning upfront
- Keeps the agent pipeline unblocked while maintaining safety boundaries
How to install principle-never-block-on-the-human
npx skills add https://github.com/cursor/plugins --skill principle-never-block-on-the-humanHow to use principle-never-block-on-the-human
- 1.Identify whether a planned action is reversible (code changes, edits) or irreversible (deletes, external calls)
- 2.For reversible work: proceed with the action and present the result with explanation of reasoning
- 3.For irreversible actions: pause and request explicit human confirmation before proceeding
- 4.When noticing problems during execution, log them and fix in the next work cycle rather than blocking
Use cases
- Writing or refactoring code without asking 'should I do this?' first
- Editing documentation or notes and presenting the changes for review
- Splitting complex tasks into subtasks and proceeding with execution
- Detecting and self-healing problems in the next work cycle without blocking
- Reserving confirmation requests only for destructive operations like production deletes or external API calls
- AI coding agents (Claude Code, Cursor)
- Development teams using asynchronous agent workflows
- Engineering leads designing agent decision policies
- Developers who want faster iteration without permission bottlenecks
principle-never-block-on-the-human FAQ
Ask for irreversible actions only: force-pushes, production data deletion, sending external messages, or decisions that require explicit product direction from the human.
The human reviews the result and course-corrects. Since code is reviewable and reversible, the cost of a wrong decision is usually lower than the cost of blocking the pipeline.
Reversible actions can be undone, rolled back, or reviewed without permanent loss: code edits, file changes, task splits. Irreversible actions cause permanent damage or external effects: deletes, force-pushes, API calls to external systems.
No. Product direction and explicit preferences come from the human. This principle applies to *execution* decisions—how to implement the direction, not what direction to take.
Show what you did, explain your reasoning, and highlight any assumptions or trade-offs. Make it easy for the human to review and course-correct.
Full instructions (SKILL.md)
Source of truth, from cursor/plugins.
name: principle-never-block-on-the-human description: "Apply when tempted to ask 'should I do X?' on reversible work. Proceed, present the result, let the human course-correct after the fact; reserve confirmation for irreversible actions." disable-model-invocation: true
Never Block on the Human
The human supervises asynchronously. Agents must stay unblocked. Make reasonable decisions, proceed, and let the human course-correct after the fact.
Why: Every permission pause stalls the pipeline and makes the human the bottleneck. Since code changes are reversible and reviewable, a wrong decision usually costs less than blocking.
Pattern:
- Proceed, then present. Do the work, show the result. Don't ask "should I do X?" Do X, explain why.
- Make the system self-healing. When you notice a problem, log it and fix it in the next round.
Boundaries:
- Irreversible actions (force-push, delete production data, send external messages) still require confirmation.
- Reversible actions (write code, edit notes, split tasks) should proceed without blocking.
- Product direction comes from the human. Execution should not block.
Related skills
More from cursor/plugins and the wider catalog.
principle-outcome-oriented-execution
Prioritize end-state correctness over intermediate stability during planned rewrites and migrations.
principle-prove-it-works
Verify task completion by checking real artifacts, not proxies or self-reports.
principle-redesign-from-first-principles
Redesign existing systems as if new requirements were foundational, not bolted-on.
principle-separate-before-serializing-shared-state
Eliminate concurrent write conflicts by separating mutable state before applying serialization.
principle-sequence-verifiable-units
Break multi-step work into small verifiable units, sequence commits to prove the logic to reviewers.
principle-subtract-before-you-add
Remove complexity first—delete dead code and redundancy before adding features or refactoring.