agentsclimarketplace

Spec acceptance harness

Skill andrey-learning-machines/swe-harness/plugins/swe-harness/skills/spec-acceptance-harness

Portable SWE harness plugin for Codex and Claude Code

Install
npx -y skills add andrey-learning-machines/swe-harness --skill spec-acceptance-harness

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Turn Specification Kit (Spec Kit) feature specifications or product requirements into executable feature acceptance coverage using Gherkin feature files, Playwright browser tests, and a story matrix that maps stories to required backend or service routes. Use when a project needs durable behavior traceability from specification to implementation proof.

SKILL.md

3.8 KB, as published. Nobody here has run it

Spec Acceptance Harness

Use this skill when a feature needs traceability from product intent to executable proof.

This workflow connects four layers:

  • Specification: a durable feature document, often from Specification Kit (Spec Kit).
  • Gherkin: a business-readable .feature file with Given, When, and Then scenarios.
  • Feature acceptance tests: browser or service checks that prove the behavior from a user perspective. Use Playwright when the behavior is visible in a web user interface.
  • Story matrix: a small route or capability ledger, commonly named story_matrix.py, that fails when a required backend or service surface disappears.
flowchart LR
  A["Spec Kit feature specification"] --> B["Gherkin feature file"]
  B --> C["Acceptance test cases"]
  C --> D["Playwright browser proof"]
  C --> E["Service or route matrix proof"]
  D --> F["Release evidence"]
  E --> F

Workflow

  1. Start from the feature specification.
    • Identify user stories, success criteria, edge cases, and non-goals.
    • Mark any unresolved product or authorization ambiguity before planning tests.
  2. Write or update the Gherkin feature file.
    • Keep scenarios business-facing.
    • Avoid implementation details in the scenario text.
    • Each scenario should describe one observable behavior.
  3. Map each scenario to proof.
    • Use Playwright for browser-visible behavior.
    • Use Application Programming Interface (API) or service tests for behavior that has no meaningful browser surface.
    • Use the story matrix for required route, event, or capability presence.
  4. Add the story matrix entry before or with implementation.
    • Record story id, required route or capability, method, owning surface, and expected authorization boundary.
    • Make missing required routes fail continuous validation.
  5. Run the feature acceptance lane.
    • Gherkin lint or parser check when available.
    • Playwright focused tests.
    • Story matrix test.
    • Relevant unit or integration tests.

Traceability Table

Use this table shape in specs, pull requests, or release notes:

StoryGherkin ScenarioPlaywright TestStory Matrix EntryStatus
S1launch assigned workflowworkflow-launch.spec.tsworkflow.launchpassing

Quality Rules

  • One feature spec can have many Gherkin scenarios.
  • One Gherkin scenario can map to one or more tests.
  • Every required route or service capability needs a matrix entry.
  • Browser acceptance tests should use user-visible locators and assertions.
  • Story matrix checks should validate capability presence and expected shape, not duplicate full integration behavior.
  • Spec Kit scripts should run from the active numbered feature branch when possible; when host branch prefixes get in the way, pass the feature explicitly through the project-supported environment variable.
  • Private project names and internal route examples must be sanitized before publishing reusable templates.

When To Stop

The acceptance harness is complete when a reviewer can answer:

  • Which user story is this behavior for?
  • Which Gherkin scenario describes it?
  • Which Playwright or service test proves it?
  • Which story matrix row protects required routes or capabilities?
  • Which command should fail if the behavior drifts?

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.