---
name: al-643-stop-playing-human-middleman-between-your-plant-and-suppl
description: "An AWS case study with Celonis shows an autonomous scheduling agent for automotive and manufacturing order-to-delivery: built with the Strands SDK on Amazon Bedrock AgentCore and connected through a Celonis MCP Server. The agent loads orders ready for partner scheduling, retrieves partner resources, checks third-party calendar APIs for conflicts, applies partner customizations, books the slot and writes the outcome back via Celonis action flows. AgentCore Identity scopes what the agent can call; Observability logs every tool call back into Celonis for review. Celonis’s Peter Hofmann frames AWS as building the agent and Celonis as providing the playbook and rules. Use when the user wants help with partner production appointment scheduling (operations) for a small business."
---

# Stop playing human middleman between your plant and suppliers: let an AI agent load ready orders, check partner calendars, pick an open slot and write the appointment back into your process system

An AWS case study with Celonis shows an autonomous scheduling agent for automotive and manufacturing order-to-delivery: built with the Strands SDK on Amazon Bedrock AgentCore and connected through a Celonis MCP Server. The agent loads orders ready for partner scheduling, retrieves partner resources, checks third-party calendar APIs for conflicts, applies partner customizations, books the slot and writes the outcome back via Celonis action flows. AgentCore Identity scopes what the agent can call; Observability logs every tool call back into Celonis for review. Celonis’s Peter Hofmann frames AWS as building the agent and Celonis as providing the playbook and rules.

## Inputs to collect first
Ask the user for anything below that they haven't already given you. Don't guess missing facts.

- Your own material (text, notes or data)

## Instructions

Do the task below with the user's inputs in place of each [bracket].

1. Write down which orders are ‘ready for partner scheduling’ and which partner systems hold availability (calendar API, shared sheet, portal).
2. Expose four tools to an agent: list ready orders, get partner capacity, propose/book a slot, write the booking back to your process system.
3. Rules: never double-book; prefer the earliest conflict-free slot within [SLA]; if no slot fits, open a human exception with the reason.
4. Run in suggest-only mode for one week—agent proposes, a planner confirms—then compare lead time and manual touches to the old process.
5. Keep an audit log of every tool call and booking; feed misses back into the rules.
6. Only then allow auto-write for a narrow order type.

## Safety checks (share these with the user)

- Check names, dates, numbers and any claim against your own source material before you use the answer. AI can misread or invent details.
- Say clearly which parts need a person to review or approve before anything is sent, paid, posted or filed.

## About this skill

- Works with: Amazon Bedrock AgentCore, Strands SDK, Celonis MCP Server, partner calendar APIs
- Based on this real-world example: https://aws.amazon.com/solutions/case-studies/celonis-agentcore-case-study/
- Free prompt from AgentPlaybook Real Estate: https://agentic-library.netlify.app/prompt/al-643-stop-playing-human-middleman-between-your-plant-and-supplier/
- To install in an agent that supports Agent Skills, save this file as SKILL.md in a folder with the same name as `name` above (for example in your skills directory). Install steps for Claude.ai, Claude Code, Cursor and Codex: https://agentic-library.netlify.app/skills/