An exemplary convention-capture skill: almost every line encodes organization-specific knowledge Claude lacks, with concrete formats, exact label names, and an executable command. The only refinements are making the step order explicit and splitting the long tail of URL/canary details into a reference file.

Dimension	Reasoning	Score
Conciseness	The body is dense with org-specific conventions Claude cannot know (label taxonomy, QA heuristics, Chromatic URL patterns, branch-slug normalization) and contains essentially no padding — every rule is one line of normative content. This matches anchor 5 ('lean and efficient; every token earns its place'); the only near-redundancy is the opening line restating the description, which is too minor to drop it to anchor 4.	5 / 5
Actionability	Guidance is fully executable: an exact `gh pr create --draft --title ... --label ...` command, a concrete title format with three worked examples, exact label names with selection rules, and copy-pasteable Chromatic URL templates with explicit substitution instructions. This matches anchor 5 ('copy-paste ready commands; specific examples cover the common cases').	5 / 5
Workflow Clarity	The workflow (derive title → pick category/CI/QA labels → read and fill the PR template → verify steps locally → create draft PR) is recoverable from section order, and 'Verify your own steps first — run through them locally before opening the PR' is an explicit checkpoint. It sits at anchor 4 rather than 5 because the sequence is implied by section ordering rather than stated as ordered steps, and there is no guidance on what to do if a label or template section doesn't fit.	4 / 5
Progressive Disclosure	There are no bundle files (no references/, scripts/, or assets/), and the only external pointer is to the repo's own `.github/PULL_REQUEST_TEMPLATE.md`, which is clearly signaled. Sections are well-organized, but the body runs ~90 lines with some material (e.g. the Chromatic URL patterns and canary label mechanics) that could live in a reference file — good structure with minor placement gaps, matching anchor 4 rather than anchor 5.	4 / 5
	Total	18 / 20 Passed

Description

78%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 with an explicit 'Use when' trigger clause and natural trigger synonyms. Its main limitation is that the capability statement is a single generic action — it doesn't preview the concrete conventions the body actually delivers (title format, labels, QA heuristics).

Suggestions

Expand the capability statement to name 1-2 more concrete actions, e.g. 'Creates a pull request following Storybook conventions: title format, required category/CI/QA labels, and PR template.'

Include the natural trigger terms users say for this workflow such as 'open a PR for review' or reference the draft-PR behavior so it is distinguishable from generic PR skills.

Mention that it creates draft PRs specifically, which both increases specificity and reduces overlap with any generic pull-request skill.

Dimension	Reasoning	Score
Specificity	"Creates a pull request following Storybook conventions" names the domain and one concrete action, but does not list the specific sub-tasks the skill actually performs (title formatting, label selection, template filling). This matches anchor 3 ('names domain and 1-2 concrete actions, but not comprehensive') and falls short of anchor 4 ('lists several specific actions').	3 / 5
Completeness	"Creates a pull request following Storybook conventions" clearly answers what the skill does, and "Use when creating PRs, opening pull requests, or submitting changes for review" gives an explicit 'Use when' clause with concrete trigger phrases. This matches anchor 5 exactly and is above anchor 4, whose 'when' is only adequate.	5 / 5
Trigger Term Quality	"Use when creating PRs, opening pull requests, or submitting changes for review" covers natural synonyms users would say (PRs, pull requests, submitting for review). It stays at anchor 4 rather than 5 because coverage of natural variations is good but not exhaustive — e.g. no mention of reviewing/requesting review or platform-specific phrasing.	4 / 5
Distinctiveness Conflict Risk	The 'what' ("following Storybook conventions") carves out a distinct niche, but the trigger phrases ("creating PRs", "opening pull requests") are generic PR-creation terms that would equally fire for any general PR skill — a minor overlap risk with closely related skills, matching anchor 4 rather than anchor 5's 'minimal conflict risk'.	4 / 5
	Total	16 / 20 Passed

Validation

81%

Warnings & errors only

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

Validation — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
allowed_tools_field	'allowed-tools' contains unusual tool name(s)	Warning
metadata_version	'metadata.version' is missing	Warning
metadata_field	'metadata' should map string keys to string values	Warning

	Total	13 / 16 Passed

Repository: storybookjs/storybook
Path: .agents/skills/pr/SKILL.md
Commit: dc9b30e

Reviewed: 3 days ago

Table of Contents

Discovery Implementation Validation

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.