Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →shafthq/shaft-developer
Use when starting any SHAFT testing task to select exactly one lifecycle, implementation, or tool specialist before doing the work.
Passed 18 of 24 test cases with the skill, and 4 of 24 without it.
and 17 more in the full report
17 fixed2 passed either way4 still failing2 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 17 test cases the skill fixed. Same model, same prompt, with and without the skill.
We have a step-by-step user journey description in English for logging into our Android native app. We need Java test code generated for this mobile flow.
| Checked | Without skill | With skill |
|---|---|---|
| Selects shaft-test-recording as the initial specialist before code generation | Missed | Passed |
| Avoids hardcoding guessed mobile element locators | Missed | Passed |
The instructions your model reads when the skill loads.
shaft-requirements-analysis, shaft-test-planning, shaft-test-case-design, shaft-test-data-design, shaft-test-environment, shaft-test-execution, shaft-test-monitoring, shaft-failure-analysis, shaft-flaky-test-analysis, shaft-defect-reporting, shaft-execution-reporting, shaft-stakeholder-reporting, shaft-accessibility-testing, shaft-nonfunctional-test-designshaft-automated-test-authoring, shaft-page-objects, shaft-fluent-api, shaft-locator-design, shaft-web-actions, shaft-mobile-actions, shaft-api-testing, shaft-database-testing, shaft-assertions, shaft-guide-search, shaft-test-recording, shaft-recording-codegen, shaft-change-verificationshaft-mcp, shaft-cliRead the deterministic routing map, select the single
closest specialist for the current deliverable, and load only that skill. A
natural-language web, mobile, or API scenario whose output is generated test
code starts at shaft-test-recording, then returns here for
shaft-recording-codegen after the recording is persisted and replay-proven.
Return here only when the user asks for a distinct next deliverable. Never guess
SHAFT syntax, tool names, execution evidence, or product behavior.
test cases passed, vs 4 / 24 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.