CtrlK
BlogDocsLog inGet started
Tessl Logo

foundation-models-on-device

Apple FoundationModels framework for on-device LLM — text generation, guided generation with @Generable, tool calling, and snapshot streaming in iOS 26+. Use when adding on-device LLM features with Apple FoundationModels on iOS 26+.

68

Quality

83%

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

82%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-organized, highly actionable reference: all five core patterns are demonstrated with executable Swift and explicit error handling. The main improvements are trimming duplicated guidance sections and moving the more detailed examples into a references/ file for better progressive disclosure.

Suggestions

Merge the redundant "When to Activate" and "When to Use" sections into one, and state the availability-check rule once instead of repeating it across Best Practices and Anti-Patterns.

Move detailed material (e.g., the full SwiftUI streaming view and the @Guide constraint catalog) into a references/ file, keeping SKILL.md as a lean overview with one-level-deep pointers.

Add a short code example for GenerationOptions(temperature:) since it is recommended in Best Practices without one.

DimensionReasoningScore

Conciseness

The body is code-heavy and lean with no padding explaining known concepts, but "When to Activate" and "When to Use" are near-duplicate sections, and the availability-check warning is repeated across the pattern, Best Practices, and Anti-Patterns sections — minor trimming opportunities.

4 / 5

Actionability

Every pattern is backed by fully executable, copy-paste-ready Swift code — availability switch, single/multi-turn sessions, @Generable types, tools with error handling, snapshot streaming, and SwiftUI integration — covering the common cases comprehensively.

5 / 5

Workflow Clarity

Each pattern has clear numbered sub-steps (define type → request output; define tool → create session → handle errors) with explicit error-handling code (do/catch, .task catch). Minor gaps: no guidance on choosing between patterns, and GenerationOptions(temperature:) is mentioned without an example.

4 / 5

Progressive Disclosure

A single well-sectioned file with clear headers and no buried references, but there is no bundle at all — detailed material such as the full SwiftUI streaming view and the @Guide constraint catalog could live in a reference file to keep SKILL.md a leaner overview.

4 / 5

Total

17

/

20

Passed

Description

83%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 that clearly and concretely states both what the skill covers and when to use it, anchored to a specific framework and platform version. Its main weakness is narrow trigger coverage: adding user-natural synonyms like "Apple Intelligence" would improve both completeness and trigger term quality.

Suggestions

Add natural synonyms users would say, e.g. "Apple Intelligence", "on-device AI", "FoundationModels", to broaden trigger coverage.

Expand the 'Use when' clause to cover more trigger scenarios, e.g. "or when the user mentions Apple Intelligence, @Generable, or on-device AI features."

DimensionReasoningScore

Specificity

The description lists four concrete capability areas — "text generation, guided generation with @Generable, tool calling, and snapshot streaming" — giving comprehensive, specific coverage of the framework's features with no vague filler.

5 / 5

Completeness

Both what ("text generation, guided generation with @Generable, tool calling, and snapshot streaming") and an explicit when ("Use when adding on-device LLM features with Apple FoundationModels on iOS 26+") are present, but the when-clause covers only a single trigger scenario rather than concrete trigger phrases users might actually say.

4 / 5

Trigger Term Quality

Natural terms like "on-device LLM", "tool calling", and "Apple FoundationModels" are present, but common user synonyms such as "Apple Intelligence" (which the body itself identifies as the user-facing feature name) and "on-device AI" are missing.

4 / 5

Distinctiveness Conflict Risk

It names a specific Apple framework with distinct niche triggers ("Apple FoundationModels", "on-device", "iOS 26+"), making it clearly distinguishable from generic cloud-LLM skills with minimal conflict risk.

5 / 5

Total

18

/

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
affaan-m/ECC
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.