execute-approved-actions

Execute approved actions via MCP servers and log results to audit logs.

Updated Jan 9, 2026
One-click install
npx skills add https://github.com/Hamza123545/Hackathon-0_Ai-Employee --skill execute-approved-actions
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: execute-approved-actions
Source: https://github.com/Hamza123545/Hackathon-0_Ai-Employee/tree/main/.claude/skills/execute-approved-actions
Command: npx skills add https://github.com/Hamza123545/Hackathon-0_Ai-Employee --skill execute-approved-actions

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill automates the secure execution of externally approved actions by reading approval files from the /Approved folder, routing to the correct MCP server, performing actions (email, social media, and browser automation), logging all activity to audit logs, and updating system status to reflect outcomes.

Core Features & Use Cases

  • Reads approval_request files from /Approved, validates expiration, required parameters, and related plan references.
  • Routes actions to the correct MCP server (email, linkedin, browser automation, etc.) and executes deterministic tasks via MCP tools.
  • Logs every execution to the mandatory audit log and updates the approval and plan statuses, with proper error handling and recovery.

Quick Start

Place a properly formatted approval file in /Approved; on the next orchestrator cycle, the skill validates, executes via MCP as configured, logs the result, and moves the file to /Done/ or /Rejected/ accordingly.

Frequently Asked Questions about execute-approved-actions

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

FAQPage Schema
How do I automate the execution of approved external actions using MCP servers?

Automating approved external actions requires placing a formatted approval file in the /Approved folder. The system validates the request, routes it to the correct MCP server (email, LinkedIn, or browser), executes the task, and logs the result to the audit log.

How does human-in-the-loop (HITL) approval routing work for browser automation and email tasks?

Human-in-the-loop approval routing works by reading approval_request files from the /Approved directory. The skill validates the parameters, selects the correct MCP server for the task, and executes the browser automation or email action deterministically.

Can I use this skill to execute approved actions on LinkedIn and other social media platforms?

Yes, you can execute approved actions on LinkedIn. The skill routes validated approval requests to the appropriate MCP server, including LinkedIn, email, and browser automation, then updates the file destination and system state upon completion.

What happens when an approval request expires or fails validation before MCP execution?

When an approval request expires or fails validation, the skill handles the error and moves the file to the /Rejected/ directory. Proper error handling and recovery ensure the system state and audit log accurately reflect the rejected execution.

Do I need to manually configure audit logging when automating tasks through MCP tools?

No, manual audit logging configuration is not needed. The skill automatically logs every execution to the mandatory audit log while updating the approval and plan statuses, ensuring all automated email and browser actions are securely recorded.

How do I trigger the execution of externally approved actions in an automated workflow?

You trigger execution by placing a properly formatted approval file in the /Approved folder. On the next orchestrator cycle, the system automatically validates the request, executes the action via the configured MCP server, and moves the file to /Done/ or /Rejected/.