kortix-executor

Execute third-party API actions from Kortix sessions without raw credentials.

Updated Jun 19, 2026
One-click install
npx skills add https://github.com/MuhammadUA/Axe --skill kortix-executor-muhammadua
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: kortix-executor
Source: https://github.com/MuhammadUA/Axe/tree/main/.kortix/opencode/skills/kortix-executor
Command: npx skills add https://github.com/MuhammadUA/Axe --skill kortix-executor-muhammadua

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill eliminates the security risks and manual effort of interacting with third-party applications and APIs from a Kortix session, removing the need to handle raw credentials or hand-roll HTTP requests that may bypass access controls.

Core Features & Use Cases

  • Unified Integration Access: Connect to all configured third-party services (Pipedream, MCP, OpenAPI, GraphQL, HTTP) through a single set of MCP tools, no per-app SDKs or raw token management needed.
  • Safe, Guided Workflow: Progressively discover available tools by intent, inspect full input schemas before calling, and avoid errors from incorrect request formatting.
  • Use Case Example: Use this Skill to send a customer support email via Gmail, create a Stripe payment charge, or post an update to Slack without ever exposing sensitive API keys to the agent.

Quick Start

Ask the agent to send a Slack message to the #general channel with the text "Project kickoff is scheduled for Monday" to trigger an external action via the Executor.

Frequently Asked Questions about kortix-executor

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

FAQPage Schema
How do I securely connect to third-party APIs without exposing raw credentials to my agent?

Secure API access is achieved by routing requests through a unified executor that resolves secrets server-side, enforcing user access controls without exposing raw credentials to the agent.

What's the best way to automate SaaS platform actions like Slack messages or Stripe charges from an agent session?

Automating SaaS platform actions requires a unified interface that progressively discovers available integration tools by intent, allowing agents to trigger external actions like Slack messages or Stripe charges without manual HTTP requests.

How does progressive tool discovery work for external API integrations?

Progressive tool discovery for external API integrations works by allowing agents to search for available actions by intent, inspect full input schemas before calling, and avoid errors from incorrect request formatting.

Can I use MCP tools to connect to Pipedream, OpenAPI, and GraphQL services through a single interface?

MCP tools can connect to configured third-party services including Pipedream, OpenAPI, GraphQL, and HTTP applications through a single set of tools without requiring per-app SDKs or raw token management.

Do I need to manually construct HTTP requests to process payments or query internal systems from my agent?

Manual HTTP request construction is not needed to process payments or query internal systems, as the executor provides an auditable interface that automatically handles request formatting and enforces access controls.

What are the limitations of using a unified integration interface for external task scenarios?

A unified integration interface is limited to configured third-party services and requires server-side secret resolution, meaning external applications must be set up in the executor before agents can discover and call their actions.