CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

plugin-craft

Design and restructure Claude Code plugins. Use when applying the five restructure principles, adding agents, or packaging a plugin. It does not write a kit SKILL.md (that's `skill-craft`).

Quality

91%

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

92%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.

An exemplary skills-first orchestration document: concrete commands, explicit phase gates, a real validation feedback loop, and disciplined externalization of all heavy material into verified one-level-deep references. The only trimmable cost is duplicated dora/testing guidance repeated across the workflows, packaging section, and checklist.

Suggestions

Consolidate the dora review/fix/re-run instructions into the 'Packaging and Distribution' section and reduce the Quality Checklist entry to a single checkbox line pointing back to it, removing the duplicated command list.

State the local-testing workflow (`claude --plugin-dir`, `/reload-plugins`, namespaced invocation) once — in the 'Starting a brand new plugin' workflow — and reference it from the packaging blockquote and checklist instead of restating it three times.

DimensionReasoningScore

Conciseness

The body is dense, imperative, and teaches only devex-kit-specific conventions Claude would not already know, but guidance is repeated across sections: the dora review loop appears in both 'Packaging and Distribution' and the Quality Checklist, and local-testing instructions (`claude --plugin-dir`, `/reload-plugins`) appear three times (workflow step 6, the packaging blockquote, and the checklist). The body also slightly exceeds its own 200-line target. This is more than the 'every token earns its place' of a 5, but the repetition is structural recapping rather than the padded over-explanation of a 3.

4 / 5

Actionability

Guidance is fully executable: a copy-paste directory tree, exact frontmatter conventions, concrete commands (`claude --plugin-dir ./your-plugin`, `/reload-plugins`, `dora review <path> --format json`, `npx @hacksmith/doraval`, `claude plugin validate`), and four numbered workflows with literal file names like `commands/review-foo.deleted.md`. The few withheld details (frontmatter template, agent `<example>` syntax) are deliberately one load-call away in real reference files, and the common cases are covered.

5 / 5

Workflow Clarity

The lifecycle is clearly sequenced via the phase table and explicit phase-gate checkpoints between transitions, and the dora loop is a textbook validate→fix→re-validate feedback loop ('Re-run dora review ... until exit 0', 'Do not report done while the exit code is 1 or 2'). A quality checklist caps the process, matching the top anchor.

5 / 5

Progressive Disclosure

Every heavy section is delegated through a clearly signaled one-level-deep blockquote ('> For ... load `references/xxx.md`'), and all five referenced files (restructure-best-practices, plugin-structure, skill-development, agent-development, examples-and-packaging) actually exist in references/. The SKILL.md itself is the intended lean overview with a decision table, so navigation is easy and content is appropriately split.

5 / 5

Total

19

/

20

Passed

Description

87%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: third-person voice, explicit 'Use when' triggers covering the main invocation scenarios, and an explicit boundary statement against the nearest overlapping skill. The only weaknesses are a handful of missing natural synonyms ('build', 'create', 'refactor') and slightly abstract action phrasing.

DimensionReasoningScore

Specificity

The description lists several concrete actions — "Design and restructure Claude Code plugins", "applying the five restructure principles, adding agents, or packaging a plugin" — with only minor gaps (e.g., commands shims and README/packaging artifacts are only implied by "packaging"). It names more than the 1–2 actions of the midpoint anchor, but coverage is not comprehensive enough for a 5.

4 / 5

Completeness

Both questions are explicitly answered: the "what" ("Design and restructure Claude Code plugins") and the "when" ("Use when applying the five restructure principles, adding agents, or packaging a plugin") with concrete trigger phrases. This matches the anchor for clearly and explicitly answering both; a 4 would require the 'when' to be less explicit, which it is not.

5 / 5

Trigger Term Quality

Trigger phrases like "restructure", "adding agents", "packaging a plugin", and "Claude Code plugins" map well to what a user would naturally say. A few natural synonyms are missing (e.g., "build", "create", "refactor", "plugin architecture"), which keeps it below the comprehensive-synonym coverage of a 5.

4 / 5

Distinctiveness Conflict Risk

The niche (Claude Code plugin architecture) is distinct, and the description actively de-risks the closest sibling conflict with "It does not write a kit SKILL.md (that's `skill-craft`)", an explicit exclusion. Clear niche with distinct triggers and minimal conflict risk.

5 / 5

Total

18

/

20

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.

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

referenced_paths_exist

Referenced path issues: 3 missing

Warning

Total

15

/

16

Passed

Repository

saif-shines/devex-kit

Path

plugins/tooling/skills/plugin-craft/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.