Manage test cases, plans, and runs

Manage TCM work through your AI assistant.

Everything you can do in Mozark's Test Case Manager (TCM) web UI, you can ask the assistant to do.

Test cases

You say	Tool used
"Show me TC-27."	tcm_get_test_case
"List test cases in this project."	tcm_list_test_cases
"Find test cases about refund flows."	search_test_cases (semantic)
"Create a test case for … with these steps …"	tcm_create_test_case (manual authoring)
"Change TC-27's priority to High."	tcm_patch_test_case
"Delete TC-30 and TC-31."	tcm_bulk_delete_test_cases

Manual vs AI-generated. tcm_create_test_case is for cases you dictate yourself. AI-generated drafts are saved with save_test_cases so they carry provenance tags. The assistant picks the right one automatically.

Folders & requirements

"Create a folder called Payments and move TC-10 to TC-15 into it."

"Add a requirement 'Users can pay with UPI' and link TC-12 to it."

Tools: tcm_create_folder, tcm_list_folders, tcm_get_folder, tcm_add_test_cases_to_folder, tcm_remove_test_cases_from_folder, tcm_create_requirement, tcm_list_requirements, tcm_update_requirement.

Test plans

Building a plan is a guided flow:

1

List existing plans

The assistant lists existing plans.

2

Propose cases

It asks Mozark's Planner Agent (suggest_test_plan) to propose cases for your release or sprint. If the planner needs more detail ("How many test cases?"), that question is relayed to you.

3

Review matched test cases

You see the matched test cases.

4

Choose where to add them

Choose: create a new plan or add to an existing one.

5

Confirm changes

Nothing is created or changed until you confirm.

"Build a regression plan for release 3.2 covering login, checkout and refunds."

Tools: tcm_list_test_plans, tcm_get_test_plan, tcm_create_test_plan, tcm_update_test_plan, tcm_add_test_cases_to_plan, tcm_remove_test_cases_from_plan.

Test runs

"Create a test run from the Release 3.2 plan and assign it to me."

"What's the status of TR-8?"

"Mark step 3 of TC-12 in TR-8 as failed."

Runs can be executed by hand (you tell the assistant each step's result), or on a real device where the assistant drives the phone and records each step for you — see Run tests on a real device.

Tools: tcm_create_test_run, tcm_list_test_runs, tcm_get_test_run, tcm_patch_test_run, tcm_add_test_cases_to_run, tcm_test_run_start_testcase, tcm_test_run_set_step_status, tcm_test_run_add_evidence, tcm_test_run_finish_testcase, tcm_test_run_lock / unlock.

Bugs

"Log a High-severity bug 'Checkout crashes on coupon apply' linked to TC-12."

Tools: tcm_create_bug, tcm_list_bugs, tcm_get_bug, tcm_patch_bug, tcm_update_bug, tcm_delete_bug, tcm_bulk_delete_bugs.

Own-Cloud devices

If you execute on your own device fleet rather than Mozark's, register those devices so runs are still tracked:

"Register my Samsung S23 as a custom device for this project."

Tools: tcm_create_custom_device, tcm_list_custom_devices, tcm_execute_testcase_owncloud.