Skip to main content

Run QA

In brief: Run QA is the composite PR/change workflow—optional scope, policy-driven subflows, synonyms /testchimp test and /testchimp run QA.

When to use​

Prefer Run QA when you want one approved plan that covers the standard QA loop end-to-end for a change set (usually a PR). Use atomic workflows instead when you only need one step (for example Create tests alone).

Prompt​

/testchimp run QA
/testchimp run QA for checkout
/testchimp test plans/stories/billing
/testchimp run QA --policy run-qa-smoke.policy.md

Synonyms: /testchimp test, /testchimp run QA → catalog id run-qa.

Scope​

Scope is optional. When omitted:

  • Feature / PR branch — changes on that branch (diff vs merge base / default branch).
  • Default branch — changes since the last run-qa execution (get-last-run-workflow-detail), or ask for a history range when unclear.

Pass an explicit feature area, plans path, or file list to constrain the whole composite. All subflows share that one scope. See Workflows intro — Scope.

Policy​

Default file: plans/knowledge/policies/run-qa.policy.md (seeded by init).

Default Subflows:

  1. author-plans
  2. connect-to-test-env
  3. create-tests
  4. run-smart-regression
  5. run-explorechimp
  6. instrument-truecoverage

Customize order, skips, and quality bars in the policy—or invoke an alternate file with --policy. Useful for smoke vs full runs, or skipping ExploreChimp / TrueCoverage while optimizing.

Requires a usable connect-to-test-env policy (or Missing config on dependent steps).

What it does​

Plan → approve → Execute: the agent authors/updates requirement plans for the scope, connects to the test environment, writes or updates SmartTests and fixtures, runs smart regression (and helps fix failures), optionally runs ExploreChimp, and advances TrueCoverage instrumentation—reporting each mutative action with policy version and a single workflow_execution_id.

See also​

FAQ

What if I am not on a pull request?

Pass an explicit focus area—a user story, scenario folder, or feature scope—so the agent still limits work to relevant plans and tests.

Does /testchimp test run ExploreChimp?

Optionally in Phase 5 when UX analytics along SmartTest journeys are needed; use /testchimp explore when exploration is the primary goal.

How are tests linked to requirements?

SmartTests include // @Scenario: comments pointing to markdown scenarios, enabling traceability and folder-level coverage insights.

What MCP data does the agent use?

Plans, execution history, TrueCoverage gaps, open bugs, and Atlas screen-state context via @testchimp/cli MCP tools.

Related documentation