Testrail Importer

Import test cases into TestRail

This skill imports the reviewed CSV for a feature into TestRail in a single action. There are no modes — running it creates a section named after themfeature and adds the whole structure and every test case inside that section.

The importer never modifies the reviewed CSV, manifest, or review report.

Invocation

/testrail-importer <feature_name>

Example:

/testrail-importer google-lens

The importer is a thin wrapper over scripts/testrail_import.py. Run:

python3 scripts/testrail_import.py <feature_name>

Input files

Resolve paths from <feature_name>:

• Input directory: work/outputs/<feature_name>/
• CSV: work/outputs/<feature_name>/<feature_name>-testcases.csv. If that exact name is absent, use the single *.csv in the directory (ignore files starting with ~ or .). Stop with an error if zero or more than one CSV exists.

Never modify any input file.

Generated output:

• work/outputs/<feature_name>/testrail_import_report.md

Environment variables

Required:

TESTRAIL_URL
TESTRAIL_USER
TESTRAIL_API_KEY
TESTRAIL_PROJECT_ID
TESTRAIL_SUITE_ID            # target suite for Critical / High / Medium priority cases
TESTRAIL_SUITE_ID_LOW       # target suite for Low priority cases (required only when the CSV contains any Low-priority case)

Priority-based suite routing: cases with Critical, High, or Medium priority are created in TESTRAIL_SUITE_ID; cases with Low priority are created in TESTRAIL_SUITE_ID_LOW. Each suite gets its own root section + section structure. If the CSV has Low-priority cases and TESTRAIL_SUITE_ID_LOW is not set, the import fails before creating anything.

Optional (with defaults):

TESTRAIL_TEMPLATE_ID
TESTRAIL_USE_SEPARATED_STEPS   = false
TESTRAIL_EXTERNAL_ID_FIELD     = custom_external_id
TESTRAIL_PRECONDITIONS_FIELD   = custom_preconds
TESTRAIL_STEPS_FIELD           = custom_steps
TESTRAIL_EXPECTED_FIELD        = custom_expected
TESTRAIL_SEPARATED_STEPS_FIELD = custom_steps_separated
TESTRAIL_SUB_TEST_SUITE_FIELD  = custom_sub_test_suites
TESTRAIL_REFERENCES_FIELD      = refs
TESTRAIL_REQUEST_TIMEOUT       = 30
TESTRAIL_MAX_RETRIES           = 3

Sub Test Suite(s): the CSV Sub Test Suite(s) labels (e.g. Special Case, Regression, Telemetry) are mapped to the option ids of the TESTRAIL_SUB_TEST_SUITE_FIELD multi-select case field and imported. If that field does not exist in the project it is skipped; any label that isn't a valid option is skipped with a warning (the case is still created).

A local .env is loaded automatically when python-dotenv is installed. Never print the API key or write it to any file.

Steps rendering: set TESTRAIL_USE_SEPARATED_STEPS=true together with a separated-steps TESTRAIL_TEMPLATE_ID (e.g. "Test Case (Steps)") to store each step paired with its expected result. Otherwise Steps and Expected Results are written to the plain-text fields (custom_steps / custom_expected), which requires a text template.

CSV format

Eight columns (order may vary; headers matched by alias):

Section (Folder), Title, Preconditions, Steps, Expected Results, Priority, Type, Sub Test Suite(s)

• Section (Folder) may use /, >, or :: to express nested sub-sections.
• Steps and Expected Results use matching line indices (1., 2., …).
• Priority is one of Critical, High, Medium, Low (P0–P3 aliases accepted); it drives suite routing (see Environment variables).
• Sub Test Suite(s) is a comma-separated list of tags (e.g. Special Case, Regression, Telemetry).

What the importer does

1. Parse and validate the CSV (required columns present; every row has a section, title, steps, expected results, priority, and type; steps pair with expected results).
2. Connect to TestRail and read priorities, case types, case fields, templates, and existing sections.
3. Verify the configured content custom fields exist. The External ID and Sub Test Suite fields are optional — if they do not exist, they are skipped.
4. Split cases by priority: Critical/High/Medium → TESTRAIL_SUITE_ID; Low → TESTRAIL_SUITE_ID_LOW.
5. In each target suite that has cases, create a root section named after the feature (e.g. Google Lens).
6. Recreate the CSV's section structure inside that root section (per suite), reusing any sections that already exist and creating parents before children.
7. Create every test case inside its section in its routed suite, mapping Priority and Type to TestRail ids (with P0..P3 aliases and optional defaults).
8. Write testrail_import_report.md summarizing, per suite, sections created, cases created, and any failures.

The feature name is turned into the root section's display name by splitting on -/_/space and capitalizing each word (google-lens → Google Lens).

Report

work/outputs/<feature_name>/testrail_import_report.md records the target project/suite, the root section, counts (sections created, cases created, cases failed), the created case ids, and any failures. Status is COMPLETED or COMPLETED_WITH_ERRORS.

Notes

• Re-running imports again; it does not deduplicate. To reimport cleanly, remove the previously created feature section in TestRail first.
• Deletions and updates are out of scope — the importer only creates.