Skip to main content

Jama Connect Help

Use the Assistant to improve your request (Beta)

If you know the outcome you want but are unsure how to phrase the request, ask the Assistant for help.

For example, you can ask: Help me write a request to identify potentially overlapping requirements in the selected folder. Don't begin the analysis yet.

You can review and edit the suggested request before submitting it.

Example requests

Create proposals from a template — Propose five new requirements for the current Set using the Description field of SYS-102 as the template. Use the context “Android device battery-temperature risks.” Present the proposals in an editable table with columns for Name, Description, Rationale, and Source. Don't create any items yet.

Find requirements without relationships — In Folder FLD-3, list the Verification Test Cases that do not have a relationship to a requirement. Limit the result to the Items in this Folder and don’t create relationships.

Improve an unmeasurable requirement — Rewrite “The system should load quickly” as a testable requirement. Do not invent a performance threshold or operating conditions. Use labeled placeholders for values that are not available in the selected Jama Connect content.

Copy folder contents — Create copies of every Item contained in FLD-3 and place the copies in FLD-42. Do not copy the folder itself. Show the proposed operation and ask for confirmation before creating the copies.

Quick checklist

Before submitting a request, make sure you:

  • Identify the Jama Connect content the Assistant should use.

  • State the intended outcome and destination.

  • Include the fields, relationships, or constraints that matter.

  • Ask for placeholders when information is unavailable.

  • Use a focused set of content for analysis.

  • Separate recommendations from changes.

  • Review proposed work before confirming it.

These practices are not simply prompt-writing techniques. They reflect how governed engineering work is performed: with defined scope, trusted context, explicit decisions, and human review.