Copy-paste these prompts into your AI assistant and replace the [placeholders] with your actual values.
You don't need to know tool names — just describe what you want.
Open a card for its full prompt, a filled example, and what you'll get back. The more specific you are — URLs, test data, expected outcomes — the better the results.
/
All 41 prompts
No prompts match these filters. Try another search or select Clear filters.
Creating Test Cases
3 prompts
Create a new test with step-by-step user actions
Create a test case called "[Test Name]" that:
1. Goes to [URL]
2. Enters "[value]" in the [field name] field
3. Clicks the "[button name]" button
4. Verifies that [expected outcome] appears on the page
Required inputs and access
App URL, test name, actions, and expected outcome; permission to create tests in the selected workspace.
Filled example
Create a test called "Contact form" on https://example.com/contact. Enter "Demo User" in Name, enter "demo@example.com" in Email, click Send, and verify the confirmation message appears.
Expected result
A new test case with its ID and authored steps. Creating a test does not run it.
Test native mobile app flows on Android/iOS
Create a mobile test case called "[Test Name]" for [android/ios] that:
1. Opens the app
2. Taps on the "[element]" button
3. Enters "[value]" in the [field] field
4. Taps "[button]"
5. Verifies "[expected text]" is displayed
Required inputs and access
Platform, available app, and test steps; permission to create mobile tests.
Filled example
Create an Android test called "Open help" that opens our demo app, taps Help, and verifies "How can we help?" is visible.
Expected result
A mobile test case ID and steps. Select an available app and device when executing it.
Describe the flow in your own words
Create a test case for the [feature name] flow on [URL].
The user should be able to [describe the complete user journey in plain English, including what they see, click, and type].
Tag it with "[tag1, tag2]".
Required inputs and access
App URL, user journey, expected outcomes, and optional tags; test creation permission.
Filled example
Create a test for the help flow on https://example.com. Open Help, search for "shipping", and verify a shipping help article appears. Tag it "help".
Expected result
A new test case with steps and tags based on your description.
Running & Monitoring Tests
5 prompts
Execute a single test and get results
Run test case [ID] and tell me the results when it's done.
Required inputs and access
An existing test case ID and execution permission; a configured environment if the case requires one.
Filled example
Run test case 1550, share the live execution link, then monitor it until it completes and summarize the verdict.
Expected result
An execution ID and live link first, followed by the completed verdict and step results after monitoring.
Execute on a specific mobile device
Show me available [android/ios] devices and apps.
Then run test case [ID] on the [device name] device with app [app name].
Required inputs and access
Existing mobile test ID, an available app and device, and execution permission.
Filled example
List available Android apps and devices. After I select them, run test case 1550 and summarize the completed results.
Expected result
Available execution choices, followed by an execution link and final results for the selected device and app.
Wrap a suite in a plan, then execute it
Find the "[suite name]" test suite, use it in a test plan with my selected browser and environment, then execute the plan and give me the results.
Required inputs and access
Suite name, browser, environment, and permission to configure and execute a plan.
Filled example
Find the "Smoke" suite. Ask me to select a browser and environment, create a plan containing the suite, and run that plan to completion.
Expected result
A configured test plan and its execution results. Suites group cases; plans supply execution settings.
Execute a plan and follow its progress
Execute test plan [ID] and monitor its progress. Tell me when it's complete.
Required inputs and access
An existing plan ID with its execution settings and permission to run it.
Filled example
Execute test plan 42, share its live link, and monitor it until the results are available.
Expected result
A plan execution ID/link, progress, and completed case results.
See whether a run is still going or done
What's the status of test case [ID]? Is the execution done?
Required inputs and access
A test case or execution ID for a run you can access.
Filled example
Check the latest execution of test case 1550. If it is still running, show the live link; otherwise summarize the result.
Expected result
The current execution state and any available live link or completed verdict.
Generating Tests from Sources
8 prompts
Auto-generate tests that cover your code changes
Generate test cases from this code change for [app URL]:
[paste your git diff or describe the changes]
Required inputs and access
Diff text or an accessible PR, target app URL, and test creation permission. Reading a PR may require a separate repository connector.
Filled example
Generate regression tests for https://example.com from this change: the search page now shows "No matches" for an empty result. Show the proposed scenarios before creating them.
Expected result
Proposed regression scenarios and, when created, test case IDs covering the supplied change.
Generate tests from a bug or feature ticket
Generate tests from this Linear ticket for [app URL]:
Title: [ticket title]
Description: [ticket description]
Steps to reproduce: [steps if it's a bug]
Required inputs and access
Ticket text and target app URL. Retrieving a ticket directly requires access through the relevant connector.
Filled example
Generate tests for https://example.com from this ticket: Title: Contact form accepts a blank email. Expected: clicking Send with no email shows "Email is required".
Expected result
Test scenarios and case IDs derived from the ticket details.
Generate tests from your design mockups
Generate test cases from this Figma design: [Figma URL]
Required inputs and access
An accessible design URL and sufficient screen/interaction context; test creation permission.
Filled example
Generate tests from https://www.figma.com/design/demo/checkout for the checkout form. Ask for any design details you cannot access.
Expected result
Test scenarios derived from accessible design information, or a request for missing access/context.
Auto-generate API tests from your OpenAPI spec
Generate API tests from the Swagger spec at [Swagger/OpenAPI URL or file path]
Required inputs and access
An accessible OpenAPI URL or a file readable by the MCP server; target API context and creation permission.
Filled example
Generate API tests from https://example.com/openapi.json for GET /products, including a valid response and an unknown product ID.
Expected result
API test scenarios and generated test case IDs; inaccessible specs must be supplied through a supported source.
Turn a screen recording into test steps
Generate test cases from this video recording: [video URL]
Required inputs and access
An accessible recording URL and target app context; test creation permission.
Filled example
Generate test cases from https://example.com/demo/help-flow.mp4 for our help flow. Ask me to clarify any action that is unclear in the recording.
Expected result
Test cases representing the actions and outcomes that can be identified in the recording.
Import test scenarios from Excel
Generate test cases from the Excel file at [file path], sheet "[sheet name]"
Required inputs and access
A spreadsheet readable by the MCP server, sheet name, and scenario columns; creation permission. A local path is usable only when that server can read it.
Filled example
Generate cases from the "Smoke" sheet in D:\shared\scenarios.xlsx on my local MCP server. Ask me to map any ambiguous columns.
Expected result
Test case IDs generated from the supplied sheet, or a request to resolve missing source data or column mapping.
Two-phase: analyze first, then generate after your answers
Generate tests from these requirements:
[paste your requirements text, user stories, or acceptance criteria]
Required inputs and access
Requirements text or user stories, acceptance criteria, and creation permission.
Filled example
Analyze this requirement: users can search help articles and see "No matches" when none are found. Ask clarification questions before generating tests.
Expected result
Phase-one scenarios, questions, and a session ID. Generation starts after the clarification answers are supplied.
Convert n8n automation workflows to test cases
Generate ContextQA tests from this n8n workflow: [workflow JSON file path or URL]
Target app URL: [your app URL]
Required inputs and access
Workflow JSON or an accessible source and target app URL; creation permission. File paths must be readable by the MCP server.
Filled example
Generate ContextQA tests for https://example.com from this n8n workflow JSON: {"nodes":[],"connections":{}}. Tell me what information is missing before creating tests.
Expected result
Proposed coverage derived from workflow nodes and connections, or a request for enough workflow detail to proceed.
Debugging Failures
5 prompts
Full root cause analysis with telemetry
Test case [ID] failed. Investigate the failure and tell me the root cause.
Required inputs and access
A failed test case/run ID with accessible execution evidence.
Filled example
Investigate the latest failure of test case 1550. Explain the likely cause and cite the failing step and available evidence.
Expected result
Failure analysis based on available run evidence; conclusions may be limited when telemetry is missing.
See all HTTP requests/responses from the test run
Show me the network logs for the latest run of test case [ID]. Were there any failed API calls?
Required inputs and access
A test case/run ID with captured network telemetry.
Filled example
Show network logs for the latest run of test case 1550 and highlight failed requests.
Expected result
Available request/response telemetry and a summary of failed calls, if any were captured.
See browser console errors and warnings
Show me the console logs for the latest run of test case [ID]. Any JavaScript errors?
Required inputs and access
A test case/run ID with captured browser console telemetry.
Filled example
Show console logs for the latest run of test case 1550 and summarize JavaScript errors.
Expected result
Captured console entries and any errors found; an empty capture is also possible.
Watch the actual test execution recording
Get me the trace URL for test case [ID] so I can watch the recording.
Required inputs and access
A test case/run ID with an available trace or recording.
Filled example
Get the trace or recording link for the latest run of test case 1550.
Expected result
An available trace/recording URL, or an explanation that the run has no recording.
Investigate + all logs + trace in one go
Test case [ID] failed. I need the full picture:
1. Investigate the failure and root cause
2. Show me the network logs (any failed API calls?)
3. Show me the console errors
4. Get the trace URL so I can watch the recording
Required inputs and access
A failed test ID and access to its run evidence; telemetry availability varies by run.
Filled example
Investigate test case 1550 using the failing step, network logs, console errors, and trace link. Distinguish evidence from hypotheses.
Expected result
A combined failure summary and links to available evidence, with missing telemetry called out.
Searching & Exploring
4 prompts
Find tests by meaning, not just keywords
Do we have any tests that cover [feature or user flow description]?
Required inputs and access
A feature or user-flow description and access to the intended workspace.
Filled example
Do we have tests covering the contact form validation flow in this workspace?
Expected result
Relevant test cases and their IDs based on the described behavior.
Search, sort, and page through your tests
List my test cases that match "[search query]", sorted by newest first, 10 per page.
Required inputs and access
Search text, sorting preference, and page size; workspace read access.
Filled example
List test cases matching "contact", sorted newest first, five per page.
Expected result
A filtered page of test cases and available pagination information.
Read a test's metadata and authored steps
Show me the details and all steps for test case [ID].
Required inputs and access
An existing test ID in a workspace you can access.
Filled example
Show the details and all authored steps for test case 1550.
Expected result
Test metadata and its step definitions; this does not execute the test.
See the configurations you can run on
Show me all available test environments and browser/device configurations.
Required inputs and access
Workspace read access; device/app availability depends on your organization.
Filled example
List the available execution environments and browser configurations in my workspace.
Expected result
Available configuration IDs and names for choosing execution settings.
Exporting & Migrating Code
3 prompts
Generate a full Playwright TypeScript project
Export test cases [ID1, ID2, ID3] as a Playwright TypeScript file manifest, then save every returned file under [output directory].
Required inputs and access
Test case IDs and an output folder your AI client can write to.
Filled example
Export test cases 1550 and 1551 as a Playwright TypeScript manifest. Save all returned files under D:\ContextQA\exports\smoke on my computer.
Expected result
A file manifest containing paths and file contents. Your client writes those files; the MCP export does not accept an output directory.
Export to Cypress, Selenium, or other frameworks
Export test case [ID] as [playwright/cypress/selenium] code in [typescript/javascript/python/java].
Required inputs and access
Test case ID, supported framework, and language.
Filled example
Export test case 1550 as Playwright TypeScript code.
Expected result
Generated framework code returned to the client for review or saving.
Import your Cypress/Playwright tests into ContextQA
Analyze my test repo at [repo path] and tell me what framework it uses.
Then migrate those tests into ContextQA for [app URL].
Required inputs and access
A repository path readable by the MCP server, target app URL, and creation permission.
Filled example
On my local MCP server, analyze D:\ContextQA\demo-tests and propose a migration to ContextQA for https://example.com before creating cases.
Expected result
Detected framework and proposed migration; created case IDs after the source and migration choices are resolved.
Knowledge Bases
2 prompts
Custom rules that guide test generation and execution
Create a knowledge base called "[title]" with this prompt:
[Your custom instructions, e.g.: "Always check for accessibility attributes. Verify ARIA labels on all interactive elements. Test keyboard navigation."]
Required inputs and access
Title and reusable instructions; permission to create workspace knowledge bases.
Filled example
Create a knowledge base called "Help flow rules" with these instructions: verify visible confirmation messages and keyboard navigation in every generated help test.
Expected result
A knowledge base ID and the saved instructions available to guide relevant workflows.
See the knowledge bases in your workspace
Show me all available knowledge bases in our workspace.
Required inputs and access
Read access to the selected workspace.
Filled example
List the knowledge bases in my active workspace.
Expected result
Available knowledge bases, with IDs and titles.
AI Insights & Coverage
1 prompt
Activity analysis, coverage gaps, and recommendations
Give me AI insights on our test coverage. What areas need more testing? Show activity from the last [7/14/30] days.
Required inputs and access
Date range and workspace insight access; useful results require available activity data.
Filled example
Show test activity insights for the last seven days and identify potential coverage gaps.
Expected result
Activity insights and recommendations based on available data; these are not proof of complete coverage.
Workspaces, Authoring & Platform Management
9 prompts
These capabilities are also in the complete tool reference. Paths for repo scanning and asset uploads refer to the server’s filesystem; hosted servers cannot read a path on your computer.
Check your sign-in and switch workspaces
Show who I am signed in as, my organization, and my active workspace. List workspaces I can access, then switch to workspace version [ID].
Required inputs and access
An authenticated session. Workspace switching requires an accessible workspace version ID.
Filled example
Show my signed-in account, organization, and active workspace. List accessible workspaces; ask me which workspace version to select.
Expected result
Current account context and accessible workspace/version IDs. Switching happens only after the target is specified.
Preview and apply typed step edits
Read the ContextQA authoring skill. Show the supported typed actions and locators for [page]. Preview the changes needed to update test case [ID], then apply the reviewed changes.
Required inputs and access
Test ID, desired edits, locators/context, and edit permission.
Filled example
Read the authoring skill and propose typed steps to add a Help-button click to test case 1550. Show the proposed changes before applying them.
Expected result
An authoring plan using supported action and locator types, then updated test steps after review.
Search steps and copy a test across workspaces
Find steps containing [text] across this workspace. Export test case [ID] with its dependency manifest and use the copy skill to recreate it in workspace version [target ID]. Ask me for any redacted values that need reentry.
Required inputs and access
Source test/workspace and target workspace version IDs; read/create permissions. Redacted values need reentry.
Filled example
Find tests with steps containing "Help". Propose copying test case 1550 into workspace version 42 and tell me which dependencies or values need attention.
Expected result
Matching steps and a copy plan that includes dependencies and any redacted data that must be supplied.
Organize tests and bind test-data profiles
Organize test cases [IDs] in folder [name]. Show available test-data profiles and data generators, then bind profile [ID] to test case [ID].
Required inputs and access
Test case IDs, folder name, and data profile ID; appropriate organization/edit permissions.
Filled example
List data profiles for test case 1550. After I select a profile, organize the case in a "Smoke" folder and bind the selected profile.
Expected result
Available profiles and the resulting folder/profile associations after the choices are resolved.
Inspect environments, variables, and assets
List execution environments and their masked variables. Create a [name] environment with [variables], then show uploaded assets whose names contain [text].
Required inputs and access
Environment name/variables or an asset search; environment management permission for changes.
Filled example
List execution environments and their masked variables, then find uploaded assets whose names contain "invoice".
Expected result
Environment and asset metadata. Masked secrets remain masked; creating an environment is a separate requested change.
Check requirement coverage and linked tests
Show requirements for my current workspace and their coverage totals. For requirement [ID], list linked tests and uncovered cases. Search Jira defects in project [key].
Required inputs and access
Workspace requirements access and, for defects, the configured Jira integration/project key.
Filled example
Show requirement coverage totals for my workspace, then list tests linked to requirement 42 and identify uncovered requirements.
Expected result
Requirement coverage totals and linked tests based on the platform data currently available.
Review repo connections and PR impact analysis
Show provider connection status and enrolled repositories. Find the analysis for PR [number] in repository [ID], summarize its decisions and execution results, and show which items need review.
Required inputs and access
Access to provider connections/enrolled repositories, repository ID, and PR number; the PR must have available analysis.
Filled example
List enrolled repositories. For repository 12 and PR 34, summarize the recorded analysis decisions and execution results.
Expected result
Connection/repository metadata and the available PR impact analysis; missing analysis is reported explicitly.
Check credits, usage, and org defaults
Show my available credits and usage for [date range], then show organization execution defaults and onboarding progress. If I lack billing permission, explain the permission result.
Required inputs and access
Organization read access; billing information may require an additional role.
Filled example
Show my available credits and usage for September 2026, then show organization execution defaults. Explain any permission restrictions.
Expected result
Accessible credit/usage/default information, or the actual permission result when billing access is unavailable.
Load ContextQA skills before building payloads
List the ContextQA skills and read contextqa-platform, then load the skill relevant to [task] before constructing tool payloads.
Required inputs and access
A task description; no additional arguments are needed to list available skill resources.
Filled example
List ContextQA skills, read contextqa-platform, and load the authoring playbook before proposing changes to my help-flow test.
Expected result
Skill names and their guidance for the assistant; loading guidance does not execute a tool action.
Full automated reproduce → investigate → fix → verify loop
Fix this bug:
Title: [bug title]
Description: [bug description]
URL: [app URL where the bug occurs]
Steps to reproduce:
[step-by-step reproduction instructions]
Our repo is at [repo URL]
Required inputs and access
Bug details, app URL, repository access, and execution permission. Code edits and ticket posting require separate client/connector capabilities.
Filled example
Reproduce this bug on https://example.com: clicking Send with a blank email incorrectly succeeds. Our repo is https://github.com/example/demo. Investigate the run and propose a fix, then verify it and prepare a ticket report.
Expected result
A reproduction test and run evidence, followed by a fix/verification workflow. The MCP prepares the report; a separate ticket connector posts it.
Registered MCP templates
These nine templates are registered by the server. Select one in your client’s MCP prompt picker to load guidance for your assistant. Loading a template does not execute tools by itself.
For example, in Claude Code, use /mcp__contextqa__contextqa_run_test_case 1550. Replace contextqa with your configured server name and 1550 with your test case ID.