skilld
All skills
segmently-ai avatar

/segmently-test-kit

@a84bc1a
by segmently-aisegmently-ai/segmently-support-plugins1 stars

Customer-safe runtime helper contract used by Segmently Launch Assistant browser SHOW and E2E DO plans. Also trigger when Segmently Launch Assistant returns executeWith.companionSkill=segmently-test-kit, a SHOW focus/screenshot plan, or an E2E browser verification contract.

  • 4 files
  • 6.6 KB
  • Updated 3 months ago
  • GitHub

Ask your Agent

Use this Skill: https://skilld.dev/gh/segmently-ai/segmently-support-plugins/segmently-test-kit

This session only. Nothing lands on disk.

SKILL.md

≈74 tokens always: the name and description. ≈571 when used: this file. ≈1.1k more on demand in 3 files.

Segmently Test Kit

Use this skill only as the runtime companion named by Segmently Launch Assistant SHOW and E2E DO contracts. It is not a general testing framework for the customer to call directly.

When To Use

  • A segmently-launch-guide SHOW result names segmently-test-kit as the companion for browser navigation, focus, or screenshot evidence.
  • A segmently-launch-guide E2E DO result names segmently-test-kit as the companion for a generated browser driver.
  • The customer asks why browser setup, auth preflight, or verification is needed before a SHOW or E2E DO action can run.

Runtime Contract

Read references/runtime-contract.md before acting on a generated browser contract. The source of truth for the concrete action remains the JSON returned by segmently-launch-guide runners.

The helper contract is intentionally narrow:

  • Use the published segmently CLI from PATH for auth and verification reads.
  • Use playwright-bowser for observable browser control.
  • Keep SHOW read-only.
  • Run E2E DO only when the launch-guide action contract, required target ids, browser auth, and explicit execute approval are all present.
  • Never ask the customer for raw tokens, refresh tokens, service credentials, or direct database access.
  • Never require a Segmently source checkout, project build, maintainer command, hidden endpoint, or local-only helper.

Required Inputs

For SHOW, the generated contract normally needs a project, funnel, screen, and a browser base URL or editor URL. If session project context is present, do not ask for the project again; state which saved project is being used and ask only for the missing target inputs.

For E2E DO, require the action value plus every target id listed in the launch-guide contract. Do not infer a missing screen, funnel, or version from nearby prose when the runner says it is still missing.

Preflight

Follow references/browser-preflight.md before live browser work. Missing CLI auth or browser tooling is a recoverable preparation step, not a final failure.

Verification

Follow references/verification-contract.md. A mutation is complete only after the generated action executes and the named read-back verification succeeds. Dry-run output is a plan, not completion.

Source: SKILL.md on GitHub

Third-party checks

No third-party reports yet.

Provenance

Signed by skilld at a84bc1a. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub yesterday.

Activeupdated 3 months ago

README badge

README badge for segmently-ai/segmently-support-plugins/segmently-test-kit