Add visual regression reference prompt (.github/aw/visual-regression.md)

[Copilot]

• Create .github/aw/visual-regression.md as a reference prompt demonstrating visual regression testing with playwright + cache-memory
• Rewrite as a proper .github/aw/ agent prompt: simple name:/description: frontmatter, example workflow shown as a code block
• Link visual-regression.md in .github/aw/create-agentic-workflow.md at the playwright browser automation hint
• Simplify and condense instructions in visual-regression.md (144 → 59 lines)

Original prompt

---

This section details on the original issue you should resolve

<issue_title>[plan] Add visual regression reference workflow using playwright + cache-memory</issue_title>
<issue_description>## Objective

Add a reference workflow that demonstrates visual regression testing using playwright + cache-memory for baseline screenshot storage. This addresses the lowest-scoring scenario (FE-1, 3.8/5.0) from the Agent Persona Exploration discussion github/gh-aw#18250.

Context

The agent correctly suggests playwright for screenshots but leaves baseline comparison storage unaddressed. Without a reference workflow showing how to store "before" screenshots using cache-memory or artifacts, visual regression workflows are incomplete.

Approach

1. Create .github/workflows/visual-regression.md as a reference workflow that:
   • Triggers on pull_request
   • Uses playwright for capturing screenshots
   • Uses cache-memory to persist baseline screenshots between runs
   • Compares current screenshots against stored baselines
   • Uses safe-outputs to post a comment with the diff report
2. Follow the pattern of existing reference workflows like auto-triage-issues.md
3. Ensure filesystem-safe timestamp formats (no colons) in cache-memory filenames

Files to Create/Modify

• Create: .github/workflows/visual-regression.md (and recompile to generate .lock.yml)

Acceptance Criteria

• Reference workflow exists and compiles successfully
• Workflow demonstrates baseline storage with cache-memory
• Workflow posts visual diff report via safe-outputs
• Workflow follows security best practices (minimal permissions, safe-outputs for writes)
• The agentic-workflows agent can use this as a reference when asked about visual regression

Generated by Plan Command for issue #discussion #18250

• expires on Feb 27, 2026, 4:46 AM UTC

Comments on the Issue (you are @copilot in this section)

@pelikhan Create .github/aw/visual-regression.md

• Fixes [plan] Add visual regression reference workflow using playwright + cache-memory #18259

---

🔒 GitHub Advanced Security automatically protects Copilot coding agent pull requests. You can protect all pull requests by enabling Advanced Security for your repositories. Learn more about Advanced Security.

[pelikhan]

@copilot turn into a prompt

[Copilot]

@copilot turn into a prompt

Done in d96ebeb. The file is now a proper .github/aw/ prompt with simple name:/description: frontmatter — no workflow-specific YAML (on:, engine:, tools:, etc.). The body contains the complete example workflow as a code block (frontmatter + concise agent prompt) plus a brief "Key Design Decisions" section explaining the pattern choices.

[pelikhan]

@copilot link this file in the agent prompt

[Copilot]

@copilot link this file in the agent prompt

Done in abc03a6. Added a reference to .github/aw/visual-regression.md in create-agentic-workflow.md at the playwright browser automation hint (line 211), so the agent is directed to that reference whenever a visual regression testing task is detected.

[pelikhan]

@copilot make the instructions simpler , more condensed

[Copilot]

@copilot make the instructions simpler , more condensed

Done in 7c7bcf8. The file is now 59 lines (down from 144) — the agent prompt body is reduced to four short paragraphs, verbose step-by-step sections are removed, and the "Key Design Decisions" list is trimmed to the five most important points.

[Copilot]

Pull request overview

Adds a new reference prompt for visual regression testing workflows, and links it from the workflow-authoring guidance so agents can reuse the pattern when asked about screenshot comparison across PRs.

Changes:

• Added .github/aw/visual-regression.md reference prompt describing a Playwright + cache-memory baseline workflow pattern.
• Linked the new reference from .github/aw/create-agentic-workflow.md at the browser automation hint.
• Extended the dispatcher agent documentation to include a “test coverage” routing section.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

File	Description
.github/aw/visual-regression.md	Introduces a reference prompt + example frontmatter for screenshot baselines using cache-memory.
.github/aw/create-agentic-workflow.md	Adds a pointer to the new visual regression reference prompt for browser-automation use cases.
.github/agents/agentic-workflows.agent.md	Documents an additional routing option for test coverage analysis.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

The example workflow text says to use the noop safe-output when nothing changes, but safe-outputs: only configures add-comment. Add noop: to the safe-outputs config (or remove the instruction to use noop) so the workflow doesn’t fail with “safe output not enabled”.

Suggested change

	max: 1
	max: 1
	noop: {}

[Copilot]

This agent now routes “Analyzing test coverage” to .github/aw/test-coverage.md, but that prompt file doesn’t exist in the repository. Either add the missing .github/aw/test-coverage.md prompt (and ensure it’s linked consistently elsewhere), or remove/adjust this route to point at an existing prompt.

Suggested change

	### Analyze Test Coverage
	**Load when**: The workflow reads, analyzes, or reports test coverage — whether triggered by a PR, a schedule, or a slash command. Always consult this prompt before designing the coverage data strategy.
	**Prompt file**: https://github.com/github/gh-aw/blob/main/.github/aw/test-coverage.md
	**Use cases**:
	- "Create a workflow that comments coverage on PRs"
	- "Analyze coverage trends over time"
	- "Add a coverage gate that blocks PRs below a threshold"