CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

fix-design-notes

Design notes on a PR, "Design review (advisory)" in the evidence check, "worth fixing", layout.split-row, OW-LIST, a reviewer says the UI looks off. Use to read what the design review found on a PR's screenshots, fix it in code, and prove the note is gone.

Quality

98%

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

SKILL.md
Quality
Evals
Security

Quality

Content

100%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 an exemplary instruction-only skill: dense, repo-specific, and fully executable, with a numbered workflow that includes explicit validation, a re-run feedback loop, and the exception path for when a note should be kept. Content is appropriately placed for a single-file skill with clearly signaled one-level-deep references.

DimensionReasoningScore

Conciseness

Lean and dense with operational facts only — no explanations of concepts Claude already knows, and even context sentences carry triage value ("They never change the evidence verdict, but a note marked *Worth fixing* is a defect a careful designer would send back"). The brief "no longer / older PRs" clauses are current-state facts rather than date-conditional branching, so they don't warrant the time-sensitivity penalty.

5 / 5

Actionability

Fully executable throughout: a copy-paste-ready gh-api check() function, the exact reproduce command ("pnpm evals:e2e <slug> --local && pnpm --dir evals design:review -- --test-run latest --json"), and environment prerequisites ("judged notes need OPENAI_API_KEY or ANTHROPIC_API_KEY"). Placeholders like <pr> are standard parameterization, not pseudocode.

5 / 5

Workflow Clarity

The five-step "Reproduce, fix, verify" sequence has explicit validation with a feedback loop ("Re-run the same command. The note must be gone and no new *Worth fixing* note may appear") plus a human-judgment checkpoint ("Look at the screenshot yourself"), and step 5 states the CI-verified end state. Not a destructive/batch operation, so no cap applies.

5 / 5

Progressive Disclosure

A single-purpose skill with no bundle files, organized into four clear sections (read the notes, reproduce/fix/verify, the exception path, rules). References to evals/design-review/rubric.md and DESIGN.md are one level deep and clearly signaled, and nothing inlined warrants its own file.

5 / 5

Total

20

/

20

Passed

Description

95%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.

A strong description: it states what the skill does in three concrete verbs, supplies explicit and varied trigger phrases a user or reviewer would naturally produce, and occupies a distinct niche with repo-specific identifiers. The only minor gap is that action coverage is limited to the core read-fix-prove loop.

DimensionReasoningScore

Specificity

Names three concrete actions spanning the full task ("read what the design review found on a PR's screenshots, fix it in code, and prove the note is gone"), which meets "several specific actions" but stops at the core loop, and "prove the note is gone" is less concrete than the read/fix verbs.

4 / 5

Completeness

Explicitly answers both: what ("read… fix it in code, and prove the note is gone") and when (front-loaded concrete trigger phrases, equivalent to a "Use when…" clause). The when-guidance is already explicit and specific, so it exceeds anchor 4.

5 / 5

Trigger Term Quality

Comprehensive natural triggers — "Design notes on a PR", "worth fixing", "a reviewer says the UI looks off" — alongside the exact check name ("Design review (advisory)") and rule identifiers (layout.split-row, OW-LIST); no obvious synonym a user would actually say is missing.

5 / 5

Distinctiveness Conflict Risk

Clear niche with distinct repo-specific triggers (Evidence preview check, OW-LIST, layout.split-row); no plausible overlap with a generic skill, so minimal conflict risk.

5 / 5

Total

19

/

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

different-ai/openwork

Path

.opencode/skills/fix-design-notes/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.