skill CX Codex

principle-never-block-on-the-human — Claude Code skill

Q: How to install principle-never-block-on-the-human in Claude Code?

Run `npx skills add backnotprop/pstack --skill principle-never-block-on-the-human -a claude-code` in your terminal.

Instructs coding agents to execute reversible changes immediately instead of asking for approval. Useful for autonomous sessions under asynchronous supervision.

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓

npx skills add backnotprop/pstack --skill principle-never-block-on-the-human -a claude-code

Open on GitHub

What principle-never-block-on-the-human does

The skill establishes an autonomy rule: agents must remain unblocked when dealing with reversible work. Instead of asking "should I do X?", the agent executes the change and explains the reasoning afterwards. Code edits are reviewable and easy to roll back. A mistaken edit usually costs less than waiting for human intervention.

The guideline sets a strict boundary between reversible and irreversible actions. Modifying source files or breaking down tasks proceeds without pauses. Destructive actions stay gated. Pushing with --force or deleting production data still demands explicit confirmation.

The frontmatter specifies disable-model-invocation: true, so this functions as a behavioral rule rather than an automated tool script. When the agent spots a defect, it logs the problem and applies a fix in the subsequent cycle.

What it helps with

prevent agents from halting execution to ask permission for code edits
enable asynchronous agent supervision without idle waiting periods
allow the agent to log issues and apply self-healing fixes in the next pass
protect production environments by requiring confirmation for irreversible actions

Example prompts for the agent

Refactor the authentication flow following principle-never-block-on-the-human
Split the task into subtasks and proceed with code edits without waiting for approval

How to install principle-never-block-on-the-human

1Skills CLI (Vercel)

npx skills add backnotprop/pstack --skill principle-never-block-on-the-human -a claude-code

2Manually: put the files in place

~/.claude/skills/principle-never-block-on-the-human/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/principle-never-block-on-the-human/

1Skills CLI (Vercel)

npx skills add backnotprop/pstack --skill principle-never-block-on-the-human -a codex

2Manually: put the files in place

.agents/skills/principle-never-block-on-the-human/SKILL.md

For all projects use ~/.agents/skills/

1Skills CLI (Vercel)

npx skills add backnotprop/pstack --skill principle-never-block-on-the-human -a cursor

2Manually: put the files in place

.cursor/skills/principle-never-block-on-the-human/SKILL.md

Cursor also reads .agents/skills/ and .claude/skills/

1GitHub CLI

gh skill install backnotprop/pstack principle-never-block-on-the-human

2Manually: put the files in place

.github/skills/principle-never-block-on-the-human/SKILL.md

1Skills CLI (Vercel)

npx skills add backnotprop/pstack --skill principle-never-block-on-the-human -a gemini-cli

2Platform CLI

gemini skills install https://github.com/backnotprop/pstack --path skills/principle-never-block-on-the-human

Manually: put the files in place

.devin/skills/principle-never-block-on-the-human/SKILL.md

The legacy .windsurf/skills/ path still works

Skills CLI (Vercel)

npx skills add backnotprop/pstack --skill principle-never-block-on-the-human -a opencode

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about principle-never-block-on-the-human

How to install principle-never-block-on-the-human in Claude Code?

Run npx skills add backnotprop/pstack --skill principle-never-block-on-the-human -a claude-code in your terminal.

What does principle-never-block-on-the-human do?

It instructs agents to execute reversible work like code changes without stopping for permission. The human reviews results after the work finishes.

Which actions still require human confirmation?

Irreversible actions remain blocked without human approval. Examples include force-pushing branches, deleting production data, and sending external messages.

Source text

SKILL.md · MIT

---
name: principle-never-block-on-the-human
description: "Apply when tempted to ask 'should I do X?' on reversible work."
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. Since code changes are reversible and reviewable, a wrong decision usually costs less than blocking.

- 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.
- 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.

langgraph-human-in-the-loop langchain-ai Skill for building LangGraph workflows that pause and wait for human input. Use it to show agent output, request approval, or validate user responses before continuing. CX Codex +1 17,200installs messaging-the-human mvschwarz This skill covers the mechanics of an agent messaging a human through rig: filing a decision request or a quiet update, then handling the reply. CX Codex +1 ★ 8,172repo stars recipe-block-focus-time official googleworkspace Sets up recurring focus time blocks in Google Calendar to protect deep work hours. Works directly through the gws CLI tool. CX Codex +1 30,400installs setup-pstack michael-denyer Picks a model for every pstack role, detects what your session actually offers, and writes the override sheet. CX Codex +1 899installs antislop-human miqdadbadjuber An accessibility skill for AI agents that checks UI contrast, keyboard focus, and control states. Ships with a WCAG AA contrast checker script. CX Codex +1 5,900installs human-gate Alireza Rezvani Blocks an agent loop until a named person reviews the artifact and signs off on open issues. 8installs

From the same set44

principle-minimize-reader-load A skill that spots unnecessary layers of abstraction and hidden state slowing down code comprehension. Use it when reviewing complex code where the logic is hard to trace. CX Codex +1 5,200installs principle-migrate-callers-then-delete-legacy-apis Instructs the assistant to migrate internal callers and delete legacy APIs in a single wave, skipping temporary compatibility layers. CX Codex +1 5,200installs principle-model-the-domain Guides an agent to model domain rules in data structures instead of growing conditionals. Use it when stateful code or repeated shape assumptions start spreading. CX Codex +1 5,200installs principle-redesign-from-first-principles A skill that makes the agent rebuild an existing design as if the new requirement had been there from day one. Use it when a change ripples through types, docs, and examples. CX Codex +1 5,200installs principle-build-the-lever A skill that pushes the agent to write a codemod, script or generator for non-trivial work instead of editing by hand, so the result can be rerun. CX Codex +1 5,200installs principle-make-operations-idempotent A skill for writing commands and processing loops that converge to the same state regardless of restarts, crashes, or retries. Use it when operations need to survive partial runs and still reach the correct outcome. CX Codex +1 5,200installs principle-subtract-before-you-add A development skill that enforces removing dead code, stubs, and unused validators before writing new features or refactoring. CX Codex +1 5,200installs maintain-verification-skill Keep a project's verification guide and feature map aligned with its code. The skill checks each mapped feature against the live app and flags confirmed fixes. CX Codex +1 5,200installs