Overview

Two Rule Intelligence tabs prepare inputs for test generation: Data for Generating Tests turns sample spreadsheets into fixtures, and Generation Setup shows generation readiness per endpoint and lets you override how those fixture fields map into requests and assertions. Background: Understanding Rule Intelligence.

Before you begin

Open the workspace (project editor → Requirements tab, or the standalone Rule Intelligence hub → select a project), then click the Data for Generating Tests tab.
The header reads Data for Generating Tests with the note "Sample data used when generating tests. Upload Excel, CSV, or other files — their values are parsed into fixtures so generated tests use realistic inputs. (For data used when running tests, see Project Settings → Data for Running Tests / Datasets.)"
This is sample data for generating tests. To iterate a test over rows of data at run time, use Data-driven testing instead.

Step 1 — Upload sample data

On Data for Generating Tests, click Choose test data files.
Select your files. On success you get a Test data uploaded (or N test data file(s) uploaded) confirmation.
Each file lands as a row with columns Name, Type, Status, Created, and Actions. Test-data rows use simpler status labels than requirement docs: Uploaded, Ready, Parsing Queued, Parsing, Parsed, Failed.

If nothing is uploaded yet, the empty state notes "Supported: Excel, CSV, JSON, YAML, and other common data formats."

Step 2 — Extract test scenarios

For an Excel or CSV row, the Actions cell shows an Extract test scenarios from file button (table-cells icon). Click it to parse the spreadsheet into scenarios and persist them as fixtures.

On completion you get a confirmation such as "N scenario(s) extracted; N row(s) persisted for TestCompiler," and a Test Data Preview panel opens.

Step 3 — Read the Test Data Preview

The Test Data Preview panel reports:

N scenario(s) extracted.
N scenario row(s) persisted for deterministic test generation (idempotent upsert) — persisted rows feed the generator directly.
N dataset(s) created — happy-path tests will iterate these rows automatically, when datasets were produced.
A reminder: "Mapping overrides for these rows live in the Generation Setup tab."

Scenarios are grouped by sheet, one card per sheet, headed Sheet: <name> with a N rows · N columns count. Each card is a table: a # column, a Scenario column, and one column per request/response field so you can see the actual values. Re-run Extract test scenarios any time; persistence is an idempotent upsert, so it won't duplicate rows.

Step 4 — Check generation readiness

Switch to the Generation Setup tab. Its header explains: "Preview generation coverage per endpoint and (optionally) override how uploaded fixture fields map into requests, parameters, and assertions. Auto-inference applies when no override is saved."

In the Generation readiness (endpoint) card:

Pick an endpoint from the Select endpoint… dropdown.
Click Preview coverage.

The result reports, for that endpoint:

Approved requirements (project) and how many are Applicable to endpoint.
Persisted fixtures for endpoint (with a (fixture compiler flag off) note if that flag is disabled).
Mapping readiness — counts of fixtures with manual overrides, endpoint scoped, and expected status set (with a (manual override flag off) note when applicable).
Any hints the platform surfaces for improving readiness.

Only approved requirements count here — see Importing and reviewing requirements.

Step 5 — (Optional) override fixture mapping

The Optional mapping overrides card lists fixtures grouped by sheet and column shape; one mapping applies to every row in a group. Use Refresh to reload fixtures from every test-data file in the project. If there are none, you'll see "No fixtures yet — Upload an Excel file on the Data for Generating Tests tab to populate this list."

For each Sheet: <name> group, set:

Control	What it does
Linked endpoints	Scope the group to specific endpoints, or leave All endpoints (auto-match by field overlap).
Expected status	The HTTP status these rows expect. The placeholder shows the auto-inferred value from the spec.
Request mappings	For each source field, choose where it maps: query, path, header, or body.
Response assertions	For each source field, choose the assertion: fieldEquals, fieldExists, headerEquals, headerExists, or responseContains.

Every mapped field carries a source badge: auto ("Auto-inferred from API spec / data — no override saved") or override ("Override saved for this field"). Expand Show N data rows to inspect the actual values. Click Save mapping (N) to persist the group's overrides; the button shows Saving... while it runs. Auto-inference applies wherever no override is saved.

Notes

Fixtures with no explicit endpoint link are auto-matched at compile time by field overlap, so you don't have to link every group manually.
Re-extracting a file re-upserts its rows rather than duplicating them.
The Generation Setup tab is a control surface, not a document list — switching to it doesn't re-fetch the upload table.

Test Data and Generation Setup