QAEverest — AI Test Case Generation for VS Code
Generate AI-powered test cases without leaving your editor. This extension gives you two ways to work:
- QAEverest chat panel — open the QAEverest view in the Activity Bar and just chat to generate functional, API, Web (UI), Mobile, Security & Performance tests. Attach a requirements document (PDF, DOCX, XLSX, CSV, TXT, MD) or paste text, generate, and save straight into your workspace. The panel follows your VS Code theme, with a built-in light/dark toggle.
- QAEverest MCP server — registered with VS Code's built-in MCP support, so Copilot Chat (or any MCP-aware agent) can call QAEverest's tools directly.

Side panel
Connect once by pasting your API key, then use the tabs:
| Tab |
What it does |
| Test Cases |
Functional / system test cases from a document or story |
| API |
API test cases from a story or an OpenAPI / Postman spec |
| Web (UI) |
End-to-end web automation test cases (runs as a background job) |
| Mobile |
Mobile automation test cases (background job) |
| Security |
Live header / SSL / vulnerability scan of a URL, with findings |
| Performance |
Load / stress / spike / soak metrics against a URL |
- Click the QAEverest icon in the Activity Bar → Enter API key to connect (paste your
qae_ key from Admin → API Management; stored in VS Code's encrypted secret storage).
- On a generation tab, drop a document or paste text. The text is extracted locally in your editor — nothing is uploaded; only the extracted text is sent to the QAEverest API for generation.
- Set the options for the tab and Generate (or Run for Security / Performance).
- For test cases: review the cards, select the ones you want, and save to the workspace as Gherkin
.feature (BDD) or structured .json (TDD) — under qaeverest-tests/.
What you get
The same QAEverest MCP server as the Claude Code plugin and the JetBrains plugin, so Copilot Chat gets the same tools: generate test cases, understand your codebase, build and run suites, test AI agents, and run security and performance checks. Your account is the one you connected in the panel.
| Area |
Tool |
What it does |
Uses credits |
| Chat |
qae_chat |
Ask the QA copilot anything: generate tests, build or run a suite, ask about a run |
yes |
|
qae_approve |
Settle an approval qae_chat asked for, with the user's own answer |
yes |
|
qae_new_chat |
Start a fresh conversation |
no |
|
qae_history |
List this project's conversations |
no |
|
qae_open_chat |
Make an earlier conversation active again |
no |
| Generate |
qae_generate_testcases |
Functional, system, SIT, database and other test cases from a story |
yes |
|
qae_generate_api_tests |
API test cases from a description, an OpenAPI/Swagger spec or a Postman collection |
yes |
|
qae_generate_ui_tests |
End-to-end UI automation test cases (navigate / click / type / assert) |
yes |
|
qae_generate_mobile_tests |
Mobile automation test cases: gestures, device states, platform behaviour |
yes |
| Your code |
qae_scan_project |
Repo Map: languages, frameworks, entry points, files worth testing (local) |
no |
|
qae_index_project |
Index the codebase so generation can cite real code |
no |
|
qae_search_project |
Semantic search over the indexed codebase |
no |
|
qae_build_inventory |
Ranked inventory of testable features, with the files each came from |
no |
|
qae_select_features |
Re-show the inventory with the ids to generate from |
no |
|
qae_generate_from_features |
Test cases for inventory features, grounded in their source files |
yes |
| Suites and runs |
qae_list_suites |
Automation suites on the account, with their ids |
no |
|
qae_run_suite |
Execute an automation suite against an environment |
yes |
|
qae_save_report |
Write the HTML report for the last run into qaeverest-reports/ |
no |
|
qae_open_report |
Path of the most recently saved report |
no |
| AI agent testing |
qae_agent_suites |
Saved AI agent suites, the agent each tests, baseline status |
no |
|
qae_agent_run |
Run a saved AI agent suite against the agent or another endpoint |
yes |
|
qae_agent_run_status |
Progress, trust score, release gate and verdict of an agent run |
no |
| Security and performance |
qae_security_scan |
Security headers, SSL/TLS and common vulnerability checks on a URL |
yes |
|
qae_performance_test |
Load, stress, spike or soak test: latency percentiles, throughput, error rate |
yes |
| Mobile |
qae_mobile_setup |
Device farm, app binary and handset configuration (links to the web app) |
no |
| Account |
qae_status |
Connection, backend, key type, credits, requests this month, enabled services |
no |
Ready-made workflows are offered as MCP prompts too:
| Prompt |
What it does |
agent-test |
Run a saved AI agent suite and report its release-gate verdict |
chat |
Ask the QAEverest QA copilot anything — generate tests, build and run suites, or ask about a run |
generate-tests |
Generate QA test cases from a story, a file, or a feature description via QAEverest |
index-project |
Index this project into QAEverest so generated tests cite real code |
inventory |
Build (or re-show) the ranked inventory of testable features in this project |
mobile-setup |
Configure QAEverest mobile automation — device farm, app binary, and handset |
new-chat |
Start a fresh QAEverest conversation |
performance-test |
Run a load, stress, spike or soak test against a URL |
report |
Save and show the HTML execution report for the last QAEverest run |
run-suite |
Run a QAEverest automation suite against an environment |
scan-project |
Build a QAEverest Repo Map of this project — languages, frameworks, and the files worth testing |
search-project |
Semantic search over the indexed codebase — ask in plain English |
security-scan |
Run a QAEverest security scan against a URL |
usage |
Show QAEverest credits remaining, key type, and which backend this project talks to |
Ask Copilot Chat things like:
Using qaeverest, generate test cases for: "User can reset their password via emailed OTP, max 3 attempts."
Setup
- Install the extension.
- Click the QAEverest icon in the Activity Bar → Enter API key to connect (or run QAEverest: Connect Account from the Command Palette) and paste your API key (Admin → API Management in your QAEverest account). It's stored in VS Code's encrypted secret storage — never written to a settings file.
- Use the side panel, or open Copilot Chat and ask for test cases — VS Code starts the QAEverest MCP server automatically on first use, using the same key.
Self-hosted or enterprise deployment? Set qaeverest.apiUrl in your VS Code settings to your instance's API base URL before step 2.
Credits & limits
Each generation/scan/test call consumes QAEverest credits per your plan; qae_status reports what's left. UI and mobile generation (qae_generate_ui_tests, qae_generate_mobile_tests) run as background jobs and may take a few minutes — the chat response arrives once the job completes. When the agent goes through qae_chat, anything that costs credits waits for your approval first.
Links
| |