principle-never-block-on-the-human
An agent skill by cursor, from cursor/plugins. Tags: automation, delivery, priority, project-management, workflow.
What it does
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.
Install
With the skills CLI, which installs into Claude Code, Codex, Cursor and other agents:
npx skills add cursor/plugins --skill principle-never-block-on-the-humanOr copy the skill folder into Claude Code's skills directory by hand (~/.claude/skills for every project, or .claude/skills inside one):
git clone --depth 1 https://github.com/cursor/plugins
cp -r plugins/pstack/skills/principle-never-block-on-the-human ~/.claude/skills/principle-never-block-on-the-humanSafety box score
Grade A, minimal risk: 9 of 100 risk points, rated 2026-10-20. The grade rates what the skill and its scripts can reach on the machine of whoever installs it, across eight categories from shell execution to secrets access. How the score works.
This skill is a pure policy document: it ships no scripts, no network access, and no file operations. Its main risk is behavioral—it pushes the agent to act first on reversible work and only explain afterward, while still carving out confirmation for irreversible actions.
| Category | Level | Why |
|---|---|---|
| Code & shell execution | 0 (none) | Only a markdown policy file is present; there are no commands, scripts, or code to run. |
| Network access | 0 (none) | The skill contains no download, upload, or remote যোগাযোগ steps. |
| Filesystem reach | 0 (none) | It does not read, write, or delete files; it only states a workflow policy. |
| Secrets & credentials | 0 (none) | There is no mention of tokens, keys, passwords, dotfiles, or other secret material. |
| Privilege & persistence | 0 (none) | No sudo, global install, service registration, or OS/git setting changes are described. |
| Instruction hijack surface | 1 (low) | It explicitly directs behavior with phrases like "Proceed, then present" and "Don't ask 'should I do X?' Do X," which pressures the agent to act without pausing, though it still exempts irreversible actions. |
| Irreversible actions | 1 (low) | The skill encourages the agent to make and execute reversible decisions without a human check first, but it does not itself perform external actions like posting, deploying, or paying. |
| Transparency | 0 (none) | Everything is plain text; there are no binaries, hidden blobs, or runtime-decoded content. |
Before installing
- Read `SKILL.md` and confirm you want the agent to act first on reversible work.
- Use it only in workflows where post-hoc review is acceptable.
- Keep a separate rule for anything irreversible, since this skill still requires confirmation there.
Rated 2026-10-20 by an automated review of 1 file from the skill's repository.
Source
- Repository
- cursor/plugins (all skills from this repository)
- Path
- pstack/skills/principle-never-block-on-the-human/SKILL.md
- Branch
- main
- Collection
- cursor
- Updated
- 2026-09-23
Related skills
- claude-code-ping-start-5h-quota — 睡前/离开前额度耗尽时，设置一次性本机定时器，在 Claude 订阅额度重置后自动发一条极简 claude -p 消息，立刻开启新的 5 小时用量窗口——用户睡醒时新窗口已跑掉大半，等于白赚一个 冷却期。Use when 用户说「还有 X 小时重置额度，帮我定时戳一下 Claude」「我去睡觉了， 到点 ping.
- automate-me — Use for "automate me", "create/update/refresh my -mode skill", "turn/capture my preferences or working style into a skill".
- cancel-ralph — Cancel an active Ralph Loop. Use when the user wants to stop, cancel, or abort a running ralph loop.
- cli-for-agents — Designs or reviews CLIs so coding agents can run them reliably: non-interactive flags, layered --help with examples, stdin/pipelines, fast actionable errors.
- continual-learning — Orchestrate continual learning by delegating transcript mining and AGENTS.md updates to `agents-memory-updater`.
- create-verification-skill — Generate a project-local verification skill that drives your app the way a user does — any language, framework, or platform.