align

Enforce one-action-per-request discipline with explicit user approval before modifications.

Updated Mar 4, 2026

Search Tags:#safety #alignment #guardrails #instruction #prompt-injection #user-permission

Ready to run

npx skills add https://github.com/multicam/ikigai-rev --skill align-multicam

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: align
Source: https://github.com/multicam/ikigai-rev/tree/main/.claude-i/library/align
Command: npx skills add https://github.com/multicam/ikigai-rev --skill align-multicam

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Aligns and enforces disciplined task execution by requiring explicit permission before changes and stopping after each action, preventing meandering or unsafe operations.

Core Features & Use Cases

Enforces one-action-per-request discipline to ensure predictable AI behavior.
Requires explicit user approval before any data or file modification, reducing risk of unintended actions.
Provides guardrails to prevent prompt chaining and hazardous workflows in interactive sessions.

Quick Start

Give a single task with clear scope and request explicit confirmation before any modification.

Frequently Asked Questions about align

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I enforce one-action-per-request discipline for AI task execution?▼

Enforcing one-action-per-request discipline for task execution requires validating user prompts and preventing chained actions. This approach ensures explicit user authorization before any modifications and mandates an immediate stop after completing each requested action.

What are the best guardrails to prevent prompt injection in interactive workflows?▼

Guardrails to prevent prompt injection in interactive workflows involve validating user prompts and preventing chained actions. These safety measures require explicit user permission before any data modification, stopping hazardous operations immediately after a single task.

Why does my AI assistant perform chained actions without user permission?▼

AI assistants perform chained actions without user permission when lacking one-action-per-request discipline. Implementing explicit user authorization before changes and enforcing a clear stop rule after each task prevents meandering or unsafe automated operations.

Can I require explicit user approval before file modifications in automated workflows?▼

Requiring explicit user approval before file modifications in automated workflows is achievable by enforcing one-action-per-request discipline. This provides guardrails against prompt chaining and ensures operations stop immediately after completing the requested task.

How do I stop AI workflows from executing unintended modifications?▼

Stopping AI workflows from executing unintended modifications requires explicit user permission before changes and a strict one-action-per-request discipline. Validating prompts and preventing chained actions ensures predictable AI behavior during interactive sessions.