CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

ark-dashboard-testing

Test Ark Dashboard with Playwright and create PRs with screenshots. Use when testing dashboard UI, taking screenshots for PRs, or reviewing dashboard changes.

Quality

82%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

72%Weight 40%Scale 1-5

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

The body is commendably terse and mostly command-driven, covering setup, testing, and PR screenshot attachment end-to-end. Its weaknesses are operational: the copy step's relative path breaks after cd /tmp, the destructive PR-body PATCH has no read-before-write, and there are no validation checkpoints (port-forward liveness, push success) between the multi-step git and gh operations.

Suggestions

Fix the path bug: after 'cd /tmp && git clone', use an absolute path for the screenshots source (e.g., 'cp <skill_dir>/screenshots/*.png') or save screenshots to an absolute location.

Add validation checkpoints: verify the port-forward with an HTTP status check before testing, and confirm 'git push' succeeded (and the raw URL returns 200) before patching the PR body.

For the PR update, first fetch the current body ('gh api repos/<org>/<repo>/pulls/XXX --jq .body') and append the screenshot markdown rather than overwriting it with -f body.

DimensionReasoningScore

Conciseness

The body is lean and command-first with no explanation of concepts Claude already knows (port-forwarding, Playwright, git). The only trims are minor: the trailing "[Claude Code Skills Documentation](...)" link adds nothing for executing the skill, and "curl http://localhost:3000 # warm up" could be folded into a validation check. This matches anchor 4 ('Efficient; minor instances of over-explanation that could be trimmed'), not 5, because of that trailing link.

4 / 5

Actionability

Mostly executable: concrete commands for setup, cloning, copying, committing, pushing, and the PR API call, plus named MCP tools (browser_navigate, browser_take_screenshot, etc.). Not 5 because the actual testing procedure is unspecified (no indication of what to navigate to or verify on the dashboard), and the copy step has a path bug — after "cd /tmp && git clone", "cp ./screenshots/*.png" resolves to /tmp/screenshots, not the skill's screenshots directory.

4 / 5

Workflow Clarity

A rough sequence exists (setup → test → screenshot → scratch repo → update PR) but validation checkpoints are absent: no check that the port-forward is live before testing (curl is labeled 'warm up', not a check), no verification the push succeeded before referencing raw.githubusercontent.com URLs, and the PR update uses "gh api ... -X PATCH -f body=\"...\"" which overwrites the entire PR body with no pre-read. Since this includes a destructive PR-body overwrite and batch screenshot upload, the missing feedback loops cap this at anchor 3 ('Steps listed but validation gaps; sequence present but checkpoints missing or implicit').

3 / 5

Progressive Disclosure

This is a short (<50 lines) single-purpose skill with no external references needed, and it uses well-organized sections (Setup, Test with Playwright, PR Screenshots) that keep each concern separate. Per the rubric's guidance for short skills with no need for external references, this matches anchor 5 ('clear overview ... well-organized sections').

5 / 5

Total

16

/

20

Passed

Description

82%Weight 40%Scale 1-5

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 well-constructed: third-person voice, explicit what-and-when structure, and natural trigger phrases tied to a specific named target. Its only weakness is that it enumerates just two actions, leaving the description less comprehensive than the body's actual coverage.

Suggestions

Enumerate one or two more concrete actions in the description (e.g., '...upload screenshots to a scratch repo and attach them to PR bodies') to raise specificity from 2 actions to a fuller list.

Add a natural synonym trigger such as 'UI testing' or 'visual testing of the dashboard' to broaden keyword coverage for users who don't say 'testing dashboard UI'.

DimensionReasoningScore

Specificity

Quotes: "Test Ark Dashboard with Playwright and create PRs with screenshots." It names the domain (Ark Dashboard, Playwright) and exactly two concrete actions (test with Playwright, create PRs with screenshots) — matching anchor 3 ('Names domain and 1-2 concrete actions, but not comprehensive'). Not 4 because no further actions (e.g., navigating specific UI flows, uploading screenshots, updating PR bodies) are enumerated in the description.

3 / 5

Completeness

The 'what' is explicit ("Test Ark Dashboard with Playwright and create PRs with screenshots") and the 'when' is explicit with concrete trigger phrases ("Use when testing dashboard UI, taking screenshots for PRs, or reviewing dashboard changes"), matching the anchor-5 example structure (clear what + 'Use when...' with concrete triggers). Not 4 because the when-clause lists three specific triggers rather than being merely adequate.

5 / 5

Trigger Term Quality

Quotes: "testing dashboard UI", "taking screenshots for PRs", "reviewing dashboard changes" — good, natural phrases a user would say. Not 5 because common variations like "UI testing", "visual testing", "screenshot PRs", or "dashboard changes" alone are partially covered but synonyms and tool-specific terms users might use (e.g., "Playwright test", "end-to-end test") are missing.

4 / 5

Distinctiveness Conflict Risk

Quotes: "Test Ark Dashboard with Playwright", "taking screenshots for PRs" — the named target (Ark Dashboard) plus the Playwright/PR-screenshot niche gives it a clear niche with distinct triggers and minimal conflict risk with generic testing or screenshot skills. Not 4 because no closely related skill would plausibly collide with this specific dashboard + PR-screenshot combination.

5 / 5

Total

17

/

20

Passed

Validation

100%

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

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

dwmkerr/dotfiles

Path

claude/skills/ark-dashboard-testing/SKILL.md

Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.