CtrlK
BlogDocsLog inGet started
Tessl Logo

testland/visual-diff-summarizer

Builds a per-PR visual-diff summary that clusters intentional vs incidental changes across snapshots emitted by Percy, Chromatic, Playwright `toHaveScreenshot`, Storybook test-runner, and other visual testing tools - groups diffs by component / route, separates "intent-aligned with PR scope" from "cascade / regression suspect", surfaces baseline-update recommendations, and emits a single PR comment that points the reviewer at the screenshots that need actual eyes. Use when a PR has 20+ visual diffs and the reviewer needs help triaging which ones to actually open.

74

Quality

93%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Low

Low-risk findings worth noting

Overview
Quality
Evals
Security
Files

Quality

Content

85%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is a well-structured, actionable 8-step workflow with real code and an explicit verification checkpoint, plus correctly signaled one-level references to existing bundle files. Its only weakness is mild verbosity in the Overview/References prose.

Suggestions

Tighten the Overview and References sections: drop the restated 'wrong-but-consistent baseline is worse than no baseline' philosophy and the list of sibling skills (junit-xml-analysis, coverage-diff-reporter) unless they are directly invoked by a step.

Consider moving the seven-row Anti-patterns table into references/report-format.md or a dedicated anti-patterns reference so the SKILL.md overview stays lean, keeping only the most load-bearing anti-patterns inline.

DimensionReasoningScore

Conciseness

Mostly efficient — it uses lean TypeScript/Bash blocks and tables rather than prose — but the Overview and References sections restate intent-vs-diff philosophy and reference sibling skills ('junit-xml-analysis', 'coverage-diff-reporter') that add tokens without guiding execution. Not the level-3 'every token earns its place' bar.

2 / 3

Actionability

Provides concrete, executable artifacts: a copy-paste `gh pr view --json` command, TypeScript interfaces and a `classify` function, and a complete CI YAML snippet with a real action (marocchino/sticky-pull-request-comment). Not pseudocode.

3 / 3

Workflow Clarity

A clearly sequenced 8-step process with an explicit validation checkpoint in Step 7 ('Checkpoint before approving: open one sample diff per aligned cluster and confirm it matches the PR's stated intent; only then run the auto-accept command'), matching the level-3 anchor with an error-recovery feedback loop.

3 / 3

Progressive Disclosure

The body is a concise overview that signals exactly two one-level-deep references ([references/tool-outputs.md] and [references/report-format.md]); both files exist in ./references/ and hold the detailed content kept out of SKILL.md, matching the level-3 anchor.

3 / 3

Total

11

/

12

Passed

Description

100%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is concrete, trigger-rich, and answers both 'what' and 'when' with a quantified activation threshold. Third person is used throughout, consistent with the guidelines.

DimensionReasoningScore

Specificity

Lists multiple concrete actions — 'Clusters snapshots', 'separates changes that match PR intent from cascade / regression suspects', 'recommends which baselines to update', and 'emits one PR comment' — matching the level-3 anchor of several specific concrete actions.

3 / 3

Completeness

Explicitly answers both what (triage/cluster/classify/recommend/emit) and when ('Use when a PR has 20+ visual diffs / changed screenshots and the reviewer needs help deciding which to open'), matching the level-3 'what AND when with explicit triggers' anchor.

3 / 3

Trigger Term Quality

Natural terms users would say are well covered — 'visual regression diffs', 'changed screenshots', 'review one by one', 'PR' — alongside tool names (Percy, Chromatic, Playwright) that map to real reviewer vocabulary.

3 / 3

Distinctiveness Conflict Risk

A clear niche (visual-diff triage at PR scale) with a quantified trigger (20+ diffs) and a specific tool list, making it unlikely to fire for the wrong skill; distinct from general visual-regression skills.

3 / 3

Total

12

/

12

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

referenced_paths_exist

Referenced path issues: 1 missing

Warning

Total

15

/

16

Passed

Reviewed

Table of Contents