AI Copilot - Affordance for AI generated examples & behavior driven mocking #1635
ludovic-pourrat
started this conversation in
Ideas
Replies: 1 comment 6 replies
-
Please read this draft proposal that outline the discussion. 1.12.x...ludovic-pourrat:microcks:feature/ai-copilot-prompt The attached YAML Open API specification outline the use case for the At the x-microcks-ai-copilot-prompt: |
This API allows you to manage financial instruments such as stocks, bonds, and ETFs. You can create, read, update, and delete instruments using their unique ISIN identifiers.
The API supports filtering instruments by type and currency, and it provides detailed information about each instrument, including its name, type, price, currency, issue date, maturity date, and interest rate.
Instructions
- Use valid ISINs (12 alphanumeric characters, e.g., US0378331005 for equity, US912828U816 for bond), accross all operations.
- For bonds, include maturityDate and interestRate.
- Use ISO 4217 currency codes (e.g., USD, EUR).
- Ensure price is a positive float.
- Demonstrate filtering by type and currency.
- Show error cases for invalid ISIN or missing required fields. And at the operation level it highlight how to instruct about specifics (e.g. the 410 error response code for the delete operation). x-microcks-ai-copilot-prompt: |
To illustrate the already deleted ISIN use the US4592001014 ISIN.
|
Beta Was this translation helpful? Give feedback.
6 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
In the current state of the AI copilot of Microcks, there is no capability to provide a prompting guidance on the LLM about how to generate the examples.
The aim of those guidance are
The proposal is to add some extensions in the Open API specification to carry on the affordances for AI examples generation, with a
x-microcks-ai-copilot-prompt
kind of definition that would be located in theinfo
and below theoperation
.The AI copilot shall then detect those extensions and inject the prompt definition by appending both the global prompt (under
info
) and the one from the operation (underoperation
) if any.Expected validated outcome is to get better relevant examples from the domain definition and behavior driven mocking.
Thoughts ?
Beta Was this translation helpful? Give feedback.
All reactions