CtrlK
BlogDocsLog inGet started
Tessl Logo

swift-6-agent-skill

Generate, review, or refactor Swift code using modern language features from Swift 6.0 through 6.4. Use when writing new Swift code, reviewing existing code for outdated patterns, migrating from Swift 5.x to 6.x, debugging concurrency warnings, or answering questions about current Swift idioms. Covers concurrency, typed throws, noncopyable types, observation, C interop, optimization controls, testing, and more.

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

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

85%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.

A well-engineered reference skill: the body is a lean, prescriptive index with a strong decision tree, error-message-to-fix feedback loops, an explicit review checklist, and a clean one-level-deep reference bundle. The main costs are cross-section repetition of the same guidance in four places and the absence of any inline code examples, both of which trim token efficiency and in-context actionability.

Suggestions

Consolidate cross-section redundancy: items like trailing commas, raw identifiers, and `weak let` currently appear in the Decision Tree, Core Guidelines, Quick Reference table, and Review Checklist — state each once canonically and reference it elsewhere.

Add 2–3 short before/after code snippets to the body for the highest-frequency modernizations (e.g., typed throws, `Mutex` for shared state) instead of routing every example to references/MIGRATION.md.

Consider moving the full Quick Reference table into references/MIGRATION.md (which already provides comprehensive old → new mappings with code) and keeping only the top 10 rows in the body as an index.

DimensionReasoningScore

Conciseness

The body is dense and prescriptive — no explanations of concepts Claude already knows, and version-sensitive detail is properly placed in the 'Quick Reference: Old → New' deprecated-patterns table. However, items like trailing commas, raw identifiers, `weak let`, and `@c` are repeated across the Decision Tree, Core Guidelines, Quick Reference table, and Review Checklist, which is trimmable token cost. Not 3 because there is no padded explanation, only redundancy; not 5 because that redundancy is real.

4 / 5

Actionability

Directives name exact spellings, flags, and APIs (`-default-isolation MainActor`, `MemberImportVisibility`, `\(optional, default: "fallback")`, `#expect(throws:)`), each grounded in an SE number. Not 3 — the guidance is concrete and executable rather than pseudocode — but not 5 because the body contains no before/after code examples for common cases; those are all deferred to references/MIGRATION.md.

4 / 5

Workflow Clarity

The Workflow Decision Tree routes five task types in a clear sequence, the debugging section maps exact compiler error strings ('Sending value of non-Sendable type...', 'Global variable ... is not concurrency-safe') to fixes — a genuine feedback loop — and the Review Checklist provides explicit validation checkpoints ('Run through the Review Checklist systematically'). This matches the anchor-5 pattern of sequence plus validation plus error-recovery loops; no destructive or batch operations apply.

5 / 5

Progressive Disclosure

All seven bundle files exist, are one level deep (references contain no nested .md references), are cited inline with SE numbers throughout the body, and are summarized with descriptive one-liners in the References section. Content is appropriately split by version with the body acting as a navigable index — the anchor-5 structure.

5 / 5

Total

18

/

20

Passed

Description

100%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 exemplary: it states concrete capabilities in third person, provides an explicit and natural 'Use when' trigger list covering all five workflow modes, and anchors the scope to a distinct, unambiguous niche (Swift 6.0–6.4 modernization). Every dimension sits at the top anchor.

DimensionReasoningScore

Specificity

Enumerates multiple concrete actions — 'Generate, review, or refactor Swift code', 'migrating from Swift 5.x to 6.x', 'debugging concurrency warnings', 'answering questions about current Swift idioms' — with comprehensive coverage of the skill's modes. Anchor 4 would require minor coverage gaps, and none are present; the wording is third-person and free of fluff.

5 / 5

Completeness

Explicitly answers both 'what' (generate/review/refactor Swift code with modern features 6.0–6.4) and 'when' via a dedicated 'Use when' clause with concrete trigger phrases — the exact shape of the anchor-5 example. Not 4, because the 'when' is fully explicit rather than improvable.

5 / 5

Trigger Term Quality

The 'Use when' clause lists natural phrases users would actually say ('writing new Swift code', 'reviewing existing code for outdated patterns', 'migrating from Swift 5.x to 6.x', 'debugging concurrency warnings'), and the 'Covers concurrency, typed throws, noncopyable types...' list adds topic keywords. This matches the comprehensive-coverage anchor; no common phrasing of the request is missing.

5 / 5

Distinctiveness Conflict Risk

Scoped to a clear niche — modern Swift language features for versions 6.0–6.4 — with distinct, unambiguous triggers ('migrating from Swift 5.x to 6.x', 'debugging concurrency warnings') that would not fire for unrelated skills. Conflict risk is minimal.

5 / 5

Total

20

/

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: 1 missing

Warning

Total

15

/

16

Passed

Repository
pchelnikov/Swift-6-Agent-Skill
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.