Index of every page on this site: llms.txt
All skills
Run it yourself

Customer-safe Segmently browser companion for SHOW walkthroughs, screenshots, and generated E2E action contracts. Also trigger when Segmently Launch Assistant returns executeWith.skill/owningSkill=playwright-bowser, mode=show, mode=e2e, or a browserPlan/driverScript/openCommand contract.

  • 1 file
  • 4.7 KB
  • Updated 2 weeks ago
  • GitHub
Use this Skill: https://skilld.dev/gh/segmently-ai/segmently-support-plugins/playwright-bowser

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

SKILL.md

≈77 tokens for metadata: the name and description. ≈1.1k when used: this file.

Description uses 3.9% of example budget

Before choosing a Skill, your Agent reads its name and description. All available Skills share that space.

  • A shorter description leaves more room for other Skills. This entry exceeds our 1% size suggestion.

In our Claude Code example, all Skill names and descriptions share 8,000 characters. This Skill uses ≈308 characters, or 3.9%.

The 1% threshold is a size suggestion. Longer descriptions can still fit.

Your model, settings, and other Skills decide how much text your Agent can read.

Example settings and source

The example uses a 200k-token context and default Claude Code settings. The count includes the name, description, separators, and when_to_use when present. Codex also counts local file paths.

Skit's source and limits: Codex 0.160.1, Claude Code 2.1.292.

Playwright Bowser

Use this skill when a user asks to show where something is in the Segmently UI, capture a screenshot, verify visible browser state, or run a browser-backed action contract produced by Segmently Launch Assistant.

This is the public browser companion shipped in the Segmently support plugin. It is intentionally smaller than the Segmently internal QA skill with the same runtime id. It uses playwright-bowser-core for browser command syntax and uses generated launch-guide contracts for Segmently-specific actions.

Preflight

Before live browser work:

node --version
npm --version
npx --version
segmently auth status
playwright-cli --help

If Node/npm/npx are missing, follow the plugin README host prerequisites first. If Segmently auth is missing, run:

segmently auth login

If browser tooling is missing, follow playwright-bowser-core setup.

SHOW Flow

  1. Resolve the user request with segmently-launch-guide when the screen or setting is ambiguous.
  2. Resolve the exact editor URL and source IDs from the request, authorized focus or saved context. Ask only for identity still missing after inspection.
  3. Open a persistent browser session with playwright-cli.
  4. Navigate to the target without changing values.
  5. Capture a screenshot when it helps the user locate the control.
  6. Return the current URL, screenshot path, and concise next step.

Use the packaged show-runner.mjs contract when Launch Assistant provides one.

Published URL Visual Check

When segmently-cli-guide or segmently-launch-guide returns a canonical published funnel URL, open that exact URL. Do not substitute api.segmently.ai or guess a default host.

playwright-cli -s=published-funnel open https://<canonical-domain><publishedUrl> --headed --persistent

Use the visible browser as the customer walkthrough surface, then capture a screenshot or run a read-only smoke path. For paid funnels, do not enter test card details or complete checkout unless the customer explicitly approves a test purchase.

Browser-Backed DO Flow

Only mutate through an explicit generated action contract:

  • the owning launch-guide runtime must name the action;
  • the project and screen target must be covered by the user's authorization;
  • the contract must include the browser driver and verification read.

Run the packaged e2e-do-runner.mjs or the exact browser commands returned by the contract. A mutation is complete only after the verification read passes.

If auth, tool setup, target ids, or verification are missing, stop and ask for that missing input. Do not claim completion from a screenshot alone.

Theme and screen authoring review

Accept theme-authoring / animation-authoring handoffs with exact destination and the saved-state receipt. Distinguish project template, onboarding snapshot, canvas screen, editor preview and published version. Inspect preview iframes and Shadow DOM where present. A canvas tile does not prove the editor has delivered a fresh save; check identity and safely reopen/remount only after resolving dirty screen and parent drafts. Never discard or save unrelated draft work implicitly.

For visual authoring check 320, 375 and 430 px, content geometry, overflow, sticky actions, loaded images and long text. Test actual answer values, selection change, back and re-entry. Mascot and custom progress checks require a complete movement cycle plus cleanup, not one screenshot. Native Play is local replay; it does not prove a remote animation write. For library work inspect an actual authorized insertion as well as its example preview. Record exactly which states and widths were observed, and retain screenshots with target/version context.

Use one working browser session and close only owned sessions when finished. Do not turn a read-only review into publication or a purchase test.

Boundaries

  • Do not read a Segmently source repository.
  • Do not use maintainer-only route books or local debug fixtures.
  • Do not invent UI selectors or click paths.
  • Do not print credentials, cookies, browser storage secrets, or token-like values.
  • Do not perform broad destructive browser actions.

Related Skills

  • playwright-bowser-core for portable browser command syntax.
  • segmently-launch-guide for request resolution and SHOW/DO contracts.
  • segmently-cli-guide for read-back verification through public Segmently CLI.

Source: SKILL.md on GitHub

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

No third-party reports yet.

Signed by skilld at f3f9f19. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub last week.

Activeupdated 2 weeks ago

README badge

README badge for segmently-ai/segmently-support-plugins/playwright-bowser