# kinora docs > Documentation for kinora, the dashboard for browser test reports (Playwright, Vitest browser mode, Storybook) across projects and over time, with an embedded trace viewer. CI runs push their results to kinora; it tracks pass rates, trends, and flaky tests, and opens the full Playwright trace inline for any failure. - [Introduction](https://docs.kinora.dev/index.md) - [Getting started](https://docs.kinora.dev/getting-started.md) ## Guides - [Playwright reporter](https://docs.kinora.dev/guides/reporter.md) - [Vitest and Storybook](https://docs.kinora.dev/guides/vitest.md) - [CLI upload](https://docs.kinora.dev/guides/cli.md) - [PR comments](https://docs.kinora.dev/guides/pr-comments.md) - [Component testing](https://docs.kinora.dev/guides/component-testing.md) - [Flaky and broken tests](https://docs.kinora.dev/guides/flaky-tests.md) - [Environment incidents](https://docs.kinora.dev/guides/environment-incidents.md) - [Alerts](https://docs.kinora.dev/guides/alerts.md) - [MCP for coding agents](https://docs.kinora.dev/guides/mcp.md) - [Desktop app](https://docs.kinora.dev/guides/desktop.md) ## Self-hosting - [Overview](https://docs.kinora.dev/self-hosting.md) - [Configuration](https://docs.kinora.dev/self-hosting/configuration.md) - [Storage & artifacts](https://docs.kinora.dev/self-hosting/storage.md) - [Upgrading & backups](https://docs.kinora.dev/self-hosting/upgrading.md) ## Reference - [Environment variables](https://docs.kinora.dev/reference/environment.md) - [REST API](https://docs.kinora.dev/reference/api.md) ## Links - Website: https://kinora.dev - App (sign up / log in): https://app.kinora.dev - Live demo: https://demo.kinora.dev - GitHub: https://github.com/kinora-dev/kinora - Full text (every page in one file): https://docs.kinora.dev/llms-full.txt - Every page is also served as HTML without the .md suffix (e.g. https://docs.kinora.dev/guides/cli/) ## Key facts - License: FSL-1.1-MIT. The server, web, and desktop are FSL-1.1 (source-available, converts to MIT after 2 years); the reporter, Vitest reporter, CLI, core, ui, mcp, and the trace viewer are MIT. - Ingest: add @kinora/reporter to playwright.config, or upload results.json with the kinora CLI from CI. Both derive the same cross-run test identity, so history stays stable regardless of upload path. - Vitest and Storybook: add @kinora/vitest-reporter to vitest.config. It uploads browser-mode and Storybook tests (Node tests only with include: 'all'), with their Playwright traces and screenshot comparisons; Storybook stories feed the Components page. - Environment incidents: a run that failed because the app under test was down (most failures could not reach a Playwright project's baseURL, or the browser crashed) is flagged and kept out of histories, flaky scores, baselines and regression alerts. @kinora/reporter/check-base-url, a global setup, stops the run before any test when a baseURL does not answer. - Balanced shards: `kinora shard` splits a Playwright suite across CI shards by each test's median duration over the latest runs (instead of by test count) and writes one Playwright test list per shard, to run with --test-list. - Quarantine: tests quarantined in the dashboard are skipped in CI with `kinora quarantine`, which writes them as a Playwright test list to run with --test-list-invert (combines with balanced shards). - Alerts: per-project notifications on new failures / regressions via Slack, email, and webhook, with an always / on-failure / on-regression policy. - PR comments: on a GitHub pull_request run or a GitLab merge request pipeline, the reporter, Vitest reporter or CLI post a summary comment on the PR/MR (pass/fail, tests newly failing vs the base branch, link to the run). GitHub uses the CI job's GITHUB_TOKEN; GitLab uses a GITLAB_TOKEN CI variable. No credentials are stored in kinora; works self-host and cloud. prStatus / --pr-status also sets a "kinora / " commit status (GitHub PR checks, or an external job of the GitLab MR pipeline), red only for tests newly failing vs the base branch, so branch protection or "Pipelines must succeed" can gate merges on it. - MCP: @kinora/mcp exposes failures, traces, and per-test history to coding agents (Claude Code, Cursor, and others) over stdio. - Desktop: an Electron app for macOS, Windows, and Linux - a local Playwright trace viewer (no account) plus an account dashboard that can re-run a failing test locally. - Self-host: one Docker Compose (KINORA_CLOUD=false); every feature, including alerts, is unlimited. Cloud has a free tier (2,500 test results/month, 1 project, 7-day retention) and paid plans from $49/month. - API: plain REST, API-key (Bearer) auth. POST /api/v1/runs to ingest a run; GET read routes for projects, runs, failures, and per-test history.