Add vostride/agent-qa

[pranshuchittora]

Summary

Adds vostride/agent-qa as a community MCP/agent toolkit for natural-language web and mobile regression testing, CI quality gates, and test-failure triage.

The structured entry conservatively classifies its browser/mobile execution as write-capable, records approval behavior as unknown, documents operator risks, and highlights its reviewable run evidence. The README count, recent-additions table, catalog row, and YAML source are updated together.

Entry checklist

• The repo URL is public and reachable.
• The entry has a specific category.
• The risk_notes field explains what could go wrong.
• The operator_note field explains why an infrastructure operator should care.
• Labels match the observed behavior, not marketing claims.
• data/repos.yaml and README.md are both updated.

Safety checklist

• No secrets, tokens, kubeconfigs, customer data, private logs, or generated artifacts are included.
• Write-capable tools document approval gates, dry-run or preview behavior, audit/evidence output, and rollback expectations.

  The entry deliberately uses human_approval: unknown; its risk note tells operators to require approval and use disposable accounts/test tenants. The project produces reviewable run artifacts, but this PR does not claim built-in dry-run or rollback guarantees.

• Credential guidance separates read-only credentials from write-capable credentials.

  Browser/mobile test accounts are inherently write-capable. The risk note recommends scoped credentials and disposable test tenants rather than asserting a read-only credential mode.

• External dependencies are disclosed here: agent-qa can use OpenAI- or Anthropic-compatible endpoints, Gemini, or local models; hosted-provider telemetry and data handling depend on the operator's chosen provider.

Validation

python scripts/validate_repos_yaml.py              # passed: 81 entries
python scripts/sync_catalog_json.py --check        # passed: 8 skill sources
python scripts/sync_readme_counts.py --check       # passed: 81 entries / 15 sections
python -m pytest -q                                # passed: 149 tests
python scripts/run_mock_eval_scenarios.py          # passed: 7/7 scenarios

The corpus-wide network audit was not run locally; the newly added repository, documentation, and license URLs were checked directly and return HTTP 200. git diff --check also passes.

Project disclosure: this submission is for agent-qa itself. The current license is FSL-1.1-ALv2 and converts to Apache-2.0 after two years.

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud