oc-approve

Gate dangerous operations behind human approval with YAML frontmatter.

23|Updated Apr 5, 2026
One-click install
npx skills add https://github.com/reedmayhew18/ClawedBack --skill oc-approve
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: oc-approve
Source: https://github.com/reedmayhew18/ClawedBack/tree/main/.claude/skills/oc-approve
Command: npx skills add https://github.com/reedmayhew18/ClawedBack --skill oc-approve

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Gates dangerous operations with a human-in-the-loop approval to prevent unintended destructive actions triggered from chat or automation.

Core Features & Use Cases

  • Human-in-the-loop approvals: Requires explicit consent before destructive commands, file deletions, or external API calls.
  • Workflow support: Manages single gates and multi-step approval chains, including pending approvals, timeouts, and completion status.
  • Auditable records: Stores and surfaces approval decisions and descriptions to the user and operators.
  • Safe defaults: Allows non-destructive operations by default and restricts risky actions to allowed channels.

Quick Start

Queue an approval request for the dangerous operation and await the user's decision.

Frequently Asked Questions about oc-approve

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

FAQPage Schema
How do I add a human-in-the-loop approval gate for destructive shell commands?

A human-in-the-loop approval gate requires explicit user consent before executing destructive shell commands. You queue an approval request for the dangerous operation and await the user's decision, with the system managing pending approvals, denials, and expirations.

What is the best way to prevent unintended destructive git operations triggered from web-chat?

Gating destructive git operations with a human-in-the-loop approval prevents unintended actions triggered from web-chat. The system restricts risky actions to allowed channels and requires explicit consent before executing file deletions or git commands.

Can I manage multi-step approval chains with pending approvals and timeouts?

Yes, you can manage multi-step approval chains with pending approvals and timeouts. The workflow support handles single gates and sequential chains, tracking completion status, timeouts, and expirations to ensure safe execution of destructive operations.

Does this approval mechanism work for external API calls and file deletions?

The approval mechanism works for external API calls and file deletions. Designed for web-chat initiated commands and high-risk tasks, it requires human consent before any destructive file operations, external API calls, or shell commands proceed.

Do I need a specific YAML frontmatter format to configure the approval workflow?

A specific YAML frontmatter format is required to configure the approval workflow. The Skill enforces a frontmatter containing name and description fields, and supports optional scripts, references, and assets for managing the approval gate.