ishanvinaysoni-netizen/AISquare Studio AutoQA

AI-powered test generation with metadata-driven organization (flow_name, tier, area) using CrewAI + Playwright

View on GitHub

Trust Signals

Scorecard Score: not yet scored
Maintenance Recency: Activelast commit Mar 30, 2026
License: Apache 2.0

Pinned Snippet

workflow.ymlSHA-pinned

uses: ishanvinaysoni-netizen/speedathon.@895bd90f21c38c7553d34d1b77724d33690e81ad # v0.2.0

tags can be moved; commit SHAs can't. why a SHA?

Inputs

name	description	required	default
openai-api-key	OpenAI API key for test generation	yes	—
qa-github-token	GitHub token with access to AutoQA repository (required for private repos)	no	${{ github.token }}
staging-url	Staging environment URL for testing	yes	—
staging-email	Test email for staging environment login	no	test@example.com
staging-password	Test password for staging environment login	no	password123
target-repo-path	Path to the target repository (usually ".")	no	.
git-user-name	Git user name for commits	no	AutoQA Bot
git-user-email	Git user email for commits	no	rabia.tahirr@opengrowth.com
pr-body	Pull request description to parse for AutoQA metadata block (```autoqa\nflow_name: ...\ntier: ...\n```)	no	${{ github.event.pull_request.body }}
test-directory	Base directory for generated tests (will create tests/autoqa/{tier}/{area}/test_{flow_name}.py structure)	no	tests/autoqa
create-pr	Create a pull request instead of pushing directly to main branch	no	false
execution-mode	Mode of operation: "generate" (new test only), "suite" (run all tests), "all" (generate + run all), "auto-criteria" (AI generates test criteria from PR diff for developer review), "gap-driven" (uses memory coverage gaps to generate test criteria for uncovered modules), or "gap-analysis" (scans for present/missing test workflows and persists results to a SQLite database)	no	generate
action-ref	Git ref (branch, tag, or SHA) to checkout for the AutoQA action repository. Defaults to main when not provided.	no	main
auto-criteria-fallback	When true, if no autoqa block is found in the PR description, automatically generate test criteria from the PR diff (Proposal 16)	no	false
auto-criteria-mode	Auto-criteria mode: "suggest" (post for review) or "auto" (proceed if high confidence)	no	suggest
auto-criteria-threshold	Confidence score threshold (0-100) for auto-proceed mode	no	85
auto-criteria-approval	Approval mechanism: "reaction" (👍), "comment" (/autoqa approve), or "label" (autoqa:approved)	no	reaction
gap-analysis-scope	Scope for gap-analysis mode: "auto" (detect from event), "pr" (PR changed files only), or "full" (entire repo)	no	auto

Outputs

name	description
test_generated	Whether a test was successfully generated
test_file_path	Path to the generated test file (e.g., tests/autoqa/A/auth/test_user_login.py)
test_results	JSON string of test execution results
generation_metadata	JSON string of test generation metadata (includes flow_name, tier, area, etag, steps)
screenshot_path	Path to the generated screenshot
etag	ETag hash for idempotency (SHA256 of metadata + steps)
flow_name	Flow name from AutoQA metadata
tier	Test tier from AutoQA metadata (A, B, or C)
area	Test area from AutoQA metadata (optional)
error	Error message if test generation or validation failed
auto_criteria_results	JSON string of auto-criteria generation results (when using auto-criteria mode)
criteria	JSON string of generated test criteria (when using auto-criteria mode, before approval)
gap_analysis_results	JSON string of gap analysis results including present and missing workflows (when using gap-analysis mode)
dashboard_results	JSON string conforming to the AutoQA Dashboard schema (gap_analysis, summary, test_cases) for populating the dashboard UI
tests_failed	Whether tests failed during execution (true/false). When true, the action defers failure to the final step so that artifacts are still uploaded.