Requirements

The Requirements module keeps what each engagement must deliver, one requirement at a time, in an organisation-wide register. Every requirement is matched to the capability model: Lintel suggests up to three capabilities and an architect accepts or rejects each one. You can then see requirements by capability and trace each one back to drivers and goals and forward to the applications that meet it. This module needs to be switched on for your organisation.

Requirements

The register of requirements across all engagements. The counts at the top show the total, how many are Musts, how many have suggestions waiting for review, how many are not matched and how many are matched.

1. Use the filters (engagement, status, type, priority, matching, free-text search) and choose Filter.
2. Select a reference to open the requirement.
3. Use Add one, Paste a list or Import Excel or CSV to add more (contributors and above).
4. Matrix, Traceability and Download CSV give other views of the same data.

By default withdrawn requirements are hidden; change Status to All to see them. On an engagement page, a smaller Requirements panel lists the first eight and links here.

Add or edit a requirement

Use this form to write one requirement properly.

1. Choose the engagement (when adding).
2. Write the requirement itself, up to 2,000 characters.
3. Record the source, the type (Functional, Non-functional or Constraint) and the priority using MoSCoW: Must, Should, Could or Won’t this time.
4. Set the status: Draft, Agreed or Withdrawn.
5. Name the priority owner, the person who decides how important it is. Add their email so they can edit it themselves.
6. Write the acceptance criteria, how you will know it is met.

Contributors can add. Architects, the person who created it and its priority owner can edit.

A requirement

Everything about one requirement on a single page: the statement, engagement, type, priority, source, owner, acceptance criteria, matching status and how it was entered (typed, pasted, imported, from an AI interview or read from a specification).

Architects and people who may edit it see Edit and Withdraw. Withdrawing keeps the record but takes it out of the default lists. The sections below cover capability matching, proposing new capabilities and linking repository elements.

Import requirements

Load many requirements from an Excel (.xlsx) or CSV file.

1. Choose Download a template to see the columns. The first row names them; only Statement is required. The others are Source, Type, MoSCoW, Priority owner, Owner email, Acceptance criteria and Status.
2. Choose the engagement and the file (up to 4 MB).
3. Choose Import.

If any row is not valid, nothing is imported and the problems are listed, so fix them and try again.

Paste a list of requirements

The quickest way to add several. Put one requirement on each line; a leading number or bullet is removed.

1. Choose the engagement and paste the list.
2. Optionally give a source for all of them.
3. Choose a type and priority; every line gets these, and you can change each later.
4. Choose Add them.

If any line is not valid, none are added. To read a full specification document instead, raise or open an engagement and use its specification box; see Requirements specification.

Requirements by capability

A grid with requirements down the side and capabilities across the top. A cross means an architect has accepted that capability for that requirement; suggestions not yet accepted do not appear.

1. Choose an engagement and select Show.
2. Read across a row to see which capabilities a requirement lands on, and down a column to see the demand on one capability.
3. Choose Download CSV for a copy.

If the grid is empty, open a requirement and accept its suggestions first. See Capability model.

Capabilities

The matching of this requirement to the capability model. Lintel compares the words in the requirement with capability names and definitions and suggests up to three, each with the reason and a confidence of High, Medium or Low.

• Architects choose Accept or Reject on each suggestion. Others see “Awaiting an architect”.
• Add a capability yourself by entering its ID, for example CAP-0012.
• Suggest again reruns the match after you edit the text.
• Nothing in the model fits marks the requirement as having no capability, which stops it looking unmatched.

New capability proposals

When nothing in the model fits, propose a new capability for the model owner to consider. Enter a name and a short reason and choose Propose; any contributor can do this. The proposal is listed with who made it and its status. It does not change the capability model by itself: the model owner decides.

Linked repository elements

Link the requirement to the repository elements that complete its chain: drivers, goals, application services and application components.

1. Enter the element ID (for example APP-0001) and choose Link.
2. Use Remove to unlink one.

Architects only. These links feed the traceability view. See An application or technology record.

Traceability

For one engagement, a table that follows each requirement along its chain: drivers, goals, the requirement, capabilities, application services and application components. Capabilities come from accepted matches; the rest come from the elements linked on each requirement. A dash means nothing is linked at that step, which shows you where the chain is broken.

Choose an engagement, select Show, and use Download CSV to take the table into a report.