Skip to content

Test Repository

jann edited this page Sep 2, 2026 · 1 revision

Test Repository

The Test Repository is Plum's home for manual testing: a living set of suites, cases and structured steps, plus a run mode a team executes together. It sits next to automation in the same project, sharing the same members and reports.

Suites and cases

  • Suites group cases. Each has a name, a priority and a stable display ID (TS-006).
  • Cases live in a suite. Each has a title, description, priority, and a stable display ID (TC-043).
  • Display IDs never change, so links and references keep working. Move a case between suites without losing its ID or history.
  • The TS / TC prefixes are configurable in Settings → Test Cases (owner / admin). "Migrate prefix" renames every existing ID; the tags in your test code are yours to update.

Structured steps

Each step has three fields testers already use:

Field
Action what to do
Test data the input
Expected result what should happen

A case reads the same whether a person runs it or an automated test covers it. Open a case to see its steps and its full run history in a tabbed panel.

Linking automation by tag

A case is marked Automated automatically when its display ID appears as a tag in the project's test code — e.g. a .feature scenario or a Playwright test tagged @TC-043. Plum re-checks on every code change.

An automated case shows its latest automated verdict inline in a test run. Claim it and record a manual result to override that verdict — useful when the automation itself is the thing that's broken.

Test runs

A test run is a set of cases the team works through together.

  1. Build — pick any cases into the run and order them.
  2. Execute — everyone opens the same board. Claim a case, then record:
    • Pass / Fail / Blocked / Skip
    • optional notes The progress bar (3 passed · 0 failed · 1 remaining) moves for the whole team in real time.
  3. Mark Complete when done. Reopen to add more results later.

Each entry stores who executed it, when, and their name as it was at the time — see Roles and Access.

Linking a test run to an automated run

Pass a run's ID to an automated trigger (testRunId in the MCP run_tests tool or the REST endpoint) and the automated cases in that run are marked pass/fail from the results automatically.

Import / export

Settings → Test Cases → Import / Export (owner / admin) round-trips the whole repository as JSON:

  • Suites and cases match on display ID, so re-importing does not duplicate.
  • A different prefix or a missing ID gets a fresh ID on import.
  • Entries with no title are skipped.

Individual reports and suites also have their own export buttons — see Reports and Session Replay.

Clone this wiki locally