Kalokoby

Guides

How teams run QA with their agent: what to install, what to ask and what comes out. Each guide ends with a live demo run and the three install steps.

Every command, file and setting is described in Kaloko’s help, which is itself living docs published with Kaloko.

By use case

Accepting a task with an AI agent, in the pull request

The agent walks the acceptance plan on local or review, Kaloko captures every screen and e-mail, checks the criteria and puts one canvas link into the PR. The reviewer accepts or returns it there.

Read the guide →
Regression before a release: compare the run with the accepted baseline

Run the flows against staging, compare with the baseline run the team accepted, and see what moved: step statuses, criteria, evidence, pixels.

Read the guide →
SEO and landing-page checks on production, read-only, every day

A read-only run of your public pages as a visitor: headings, metadata, hreflang, Open Graph, structured data, overflow, console. You see what changed since yesterday and fix it before anyone else notices.

Read the guide →
E-mail flows: capture the message next to the screen that sent it

Sign-up confirmation, order receipt, password reset. Kaloko captures the e-mail from Mailpit or letter_opener as a step of the flow, checks its subject and links, and shows it on the canvas next to the form.

Read the guide →
Mobile apps: walk Android and iOS screens like web pages

Kaloko drives Android emulators and phones over adb and the iOS Simulator, captures each screen with its UI tree on every device you name, in light and dark mode, checks labels, touch targets and crashes, and lays the app flow out on the same canvas as your web pages.

Read the guide →
Desktop apps: Electron, macOS and Windows apps on the same canvas

Kaloko opens Electron apps through Playwright with every web check, captures macOS app windows with their accessibility tree, drives Windows apps through UI Automation, and imports screens from any other tool, so desktop flows get the same review as web and mobile.

Read the guide →
Review mode and sign-off: decide only what changed

Review a run one step at a time with the keyboard, keep approvals of unchanged steps, compare screens with swipe or onion skin and hand over a signed record of who accepted what.

Read the guide →
Recordings and traces: when a screenshot is not enough

Record a step or a whole flow on video, read what happened meanwhile (clicks, requests, console) and export the run as one walkthrough video for a client or a release note.

Read the guide →
Safari, Firefox, phones and dark mode in one run

Walk the same flow in Chromium, WebKit and Firefox, on phone, tablet and TV presets, in light and dark mode, and review every combination on one canvas.

Read the guide →
Hosted runs: Kaloko checks your pages every morning

Kaloko opens the pages of a read-only scenario on its own browsers every day or week, in every language and viewport, and shares the run as if you had walked it.

Read the guide →
A changelog with pictures and docs that keep up with the app

CHANGELOG.md and the guides in your docs folder show screens from accepted runs. When a screen changes, Kaloko says which section to rewrite.

Read the guide →
Screen library: every screen of every flow, searchable

The newest screenshot of every step across your flows in one gallery. Search by title, text, URL or step id before you design or build a screen that already exists.

Read the guide →
Design a flow with your agent, together with your team

Design a flow as live HTML with your own agent and skills. The team comments as you go, and each version stores only what changed.

Read the guide →
Design reference: build and test against the design

An accepted design version is the reference: agents build from its live files and tokens, and every run is checked for pixels, structure and tokens.

Read the guide →
Your data is yours: open formats for runs and designs

Every run and every design version leaves Kaloko as plain files: HTML, CSS, images, W3C DTCG tokens, YAML scenarios and a canvas that opens from disk.

Read the guide →
From Figma to the design reference without redrawing

Import Figma frames as design steps with their text, type and variables, make them the design reference and check every implementation run against them.

Read the guide →

By agent

QA with Claude Code: install the skill, then ask in plain words

How Claude Code runs Kaloko: the skill it installs, what it may do without asking, what it asks first, the prompts that work, and how it reads reviewer feedback and the inbox.

Read the guide →
QA with Codex: the skill in skills/kaloko and a pointer in AGENTS.md

How Codex runs Kaloko: where the skill lives, how AGENTS.md points to it, the prompts that work, and how Codex reads reviewer feedback before it continues.

Read the guide →
Kaloko in CI: every pull request preview walked and commented

One command writes the workflow: it waits for the Vercel, Netlify or Cloudflare preview, walks the steps the change affects and comments on the PR.

Read the guide →

Set up for the team

Jira, Linear, Slack and Teams: returned steps where the team works

File a returned step as a Jira or Linear issue with its screenshot and notes, link runs to issues by key, and send each project's events to its own Slack or Teams channel.

Read the guide →
Company sign-in: SAML or OIDC, SCIM and the controls around them

On Enterprise, people from your domains sign in through your identity provider, SCIM adds and removes them, and the IP allowlist and audit streaming cover the rest.

Read the guide →

Skills and tools

Skill catalog: the kaloko skill and the tools we recommend next to it

What the kaloko skill makes an agent do, what each reference covers, and which third-party skills and tools fit alongside, each with its implications: what runs where, which data leaves the machine, when not to use it.

Read the guide →

What it looks like

A real run of Kaloko on its own public pages, refreshed daily. This is the canvas your team gets.

11 of 12 steps passed · 2026-11-16Open the run in Kaloko ↗

Install once, then work through your agent

Kaloko runs where your code and your agent are. The service stores and versions the results, shows the canvas and collects approvals.

  1. Add the CLI to the project
    npm install --save-dev kaloko

    Needs Node 20 or newer. Update later with npm update kaloko.

  2. Create the config and install the skill
    npx kaloko init --agent claude --org <your-org>

    The skill is copied to .claude/skills/kaloko. npx kaloko doctor checks Chrome, the config and the token.

  3. Create your organization and a token

    Create an organization; you become its admin. The start page offers a tester token in one click, later under Settings → API tokens. Put it into the project .env:

    KALOKO_TOKEN=kal_…
    TYPESAFE_API_KEY=…   # optional: semantic evaluator

Then just ask your agent

The skill teaches your agent the whole loop: it writes the acceptance plan and the scenario from the task, walks the screens, evaluates, shares the canvas, reads what reviewers said and fixes it. You don’t type the commands; you look at the canvas.

What the agent runs (or run it yourself, e.g. in CI)

The same loop by hand:

npx kaloko start --scenario docs/tasks/TASK-123/qa/scenario.yml --env local
npx kaloko walk        # playwright steps; agent/manual steps: kaloko capture
npx kaloko evaluate
npx kaloko share --pr
npx kaloko feedback    # what reviewers said, with ids to answer