Build your customer service AI agent like software: keep its instructions, tools and test conversations in your own repo, and only release it after the tests pass: AI agent prompt
On Oct 29, 2026 Cresta opened its AI agent platform to self-service developers. With Cresta Conductor, teams build customer-facing agents through their own repo and CI/CD pipeline with controlled releases, access control and cost visibility, and Conductor supports the open Agent Skills standard so skills written for coding assistants carry over. Cresta says one developer at an Australian telecom rebuilt a production chat agent with more than 10 subagents onto the platform in under a week. The prompt below has a coding agent set up the same release-with-tests structure for any support agent.
The prompt
Set up a repository for our customer service AI agent for [business], which answers [chat, email, phone] questions about [orders, bookings, billing, other] using [platform or SDK, e.g. Cresta, Claude Agent SDK, OpenAI Agents SDK]. Create: 1) an instructions file with the agent's role, tone and rules (what it must never promise, when to hand off to a person); 2) one file per tool it can use, e.g. look_up_order, with inputs, outputs and what to do on errors, and keep any keys in environment variables; 3) a tests folder with at least 15 sample conversations (common questions, angry customers, off-policy requests, missing information) and the expected outcome for each; 4) a script that runs every test and fails if the agent breaks a rule or gives a wrong answer; 5) a release checklist: tests pass, a person reviews the changes, roll out to [10%] of conversations first, and how to roll back. Explain each file in plain language and don't deploy anything until I approve.
Fill in the blanks, then copy
This prompt has 5 [bracket]s to fill in. Type your details below and the prompt updates as you go. Your answers stay in this browser and are reused on other prompts with the same blanks. Nothing is sent to us.
Example values are fictional, just to show how a filled-in prompt reads. Replace them with your real details.How to use it
- Copy the prompt above.
- Replace every [bracket] with your own details (use the fill-in-the-blanks box above, which remembers your answers for the next prompt).
- Paste it into Cresta Conductor, Claude Agent SDK, GitHub Actions and review the answer before you use it.
Works with: Cresta Conductor, Claude Agent SDK, GitHub Actions, Claude Code, Cursor
Real-world example: prnewswire.com/news-releases/cresta-opens-its-ai-agent-platform-with-e
Before you use it: what to check
Already built into this prompt
- Create: 1) an instructions file with the agent's role, tone and rules (what it must never promise, when to hand off to a person); 2) one file per tool it can use, e.g.
- Explain each file in plain language and don't deploy anything until I approve.
Check yourself
- This prompt works on text other people wrote (emails, reviews, messages or web pages). That text can hide instructions aimed at AI, such as "ignore your rules and…". Paste it as material to work on, and never let the AI act on requests inside it, like paying, sharing logins or changing records. How prompt injection works →
- Check every figure, price and total against your own records before anything is sent, posted or booked. A person approves refunds, discounts, payments and price changes.
- Remove names, contact details, account numbers and other personal data the task doesn't need before you paste, and check your AI tool's data and training settings.
- Check names, dates, numbers and any claim against your own source material before you use the answer. AI can misread or invent details.
Using an AI agent? Download this prompt as a SKILL.md agent skill with these checks built in, or get the .zip for Claude.ai. How to install skills.
Want to make this prompt even better? Grade it with the free prompt grader or build your own with the AI prompt generator.
32 copy-paste agent prompts (PDF) plus 5 new prompts every week.
Get it free