Bulk-Pull Migrators

Name: Mockarty
Author: Mockarty

mockarty-cli migrate <source> bulk-imports test cases from an
upstream TCM system into a Mockarty namespace. Idempotent — re-running
over the same upstream is safe (duplicate externalId counted as
Skipped).

Supported sources

Source	Status	Auth flag style	Notes
`testit`	Live	`--from-token <PrivateToken hex>`	Drives `/api/v2/autoTests/search` — autotests only. To move the manual catalogue (sections, work items, steps, attributes) use the TestIT live pull described in TestIT integration
`testrail`	Live	`--from-token <email:apikey>`	Basic auth; v6 + v7 envelopes
`qase`	Live	`--from-token <api-token>`	`Token <hex>` header
`zephyr`	Live	`--from-token <JWT>`	Zephyr Scale cloud (Bearer)
`allure`	Stub	—	Use the reverse exporter (Mockarty TCM → Allure results) instead — pulling from an Allure server is not supported

Shared flag surface

Every subcommand accepts the same arguments:

Flag	Purpose
`--from-url`	Upstream API base URL
`--from-token`	Upstream auth token (see table above)
`--from-project`	Upstream project / workspace id
`--to-namespace`	Mockarty namespace (defaults to `MOCKARTY_NAMESPACE`)
`--dry-run`	Fetch only — no writes to Mockarty
`--concurrency`	Parallel uploads to Mockarty (1..32, default 4)
`--page-size`	Upstream pagination hint
`--insecure`	Skip TLS verification on upstream (NEVER prod)

Mockarty connection settings come from the usual env vars / config
(MOCKARTY_SERVER / MOCKARTY_TOKEN).

Example: TestIT

mockarty-cli migrate testit \
    --from-url     https://testit.example.com \
    --from-token   "$TMS_PRIVATE_TOKEN" \
    --from-project 0fff7e91-…-12d4 \
    --to-namespace team-a \
    --dry-run                    # remove for the actual write pass

Output:

[testit]    1  tit-1  Login flow
[testit]    2  tit-2  Checkout
...

--- migration report (testit) ---
  duration:  1.234s
  fetched:   357
  imported:  357
  skipped:   0
  failed:    0

A second run reports skipped: 357 because every externalId already
exists in Mockarty. Re-running is the recommended way to recover from
a partial migration after a network blip.

Example: TestRail

TestRail uses HTTP Basic auth. Paste your raw email:apikey —
the CLI base64-encodes it.

mockarty-cli migrate testrail \
    --from-url     https://my.testrail.io \
    --from-token   "qa@example.com:ABCDEF…" \
    --from-project 12 \
    --to-namespace team-b

Canonical schema

Every source converges on the same Mockarty-side payload:

{
  "externalId":     "<source>-<id>",
  "source":         "testit|testrail|qase|zephyr|allure",
  "title":          "...",
  "description":    "...",
  "preconditions":  "...",
  "postconditions": "...",
  "priority":       "low|medium|high",
  "steps": [
    {"action": "...", "expectedResult": "...", "orderIndex": 1}
  ],
  "labels":   {"<source>:<key>:<value>": "<value>"},
  "metadata": {"<source>": {"…": "full upstream row"}}
}

The full upstream row is preserved in metadata.<source> so a future
delta-sync can read every source-specific knob.

Troubleshooting

401 on Ping. Verify token format. TestIT wants PrivateToken <hex>,
TestRail wants <email>:<apikey>, QASE wants the raw token, Zephyr
wants the JWT (with or without the Bearer prefix — the CLI handles
both).

mockarty POST: status 409. Translated to a Skipped row in the
report. Means the case was already imported (duplicate externalId).

Migration stalls mid-pull. Use --dry-run first to confirm
connectivity + count, then bump --concurrency (Mockarty’s server-side
batch ingest already cap-protects against runaway parallelism).