agentleFS
hmj1026hmj1026/dhpk/skills/swift/SKILL.md

swift

Skill2 starsChanged 6 days agoMITView on GitHub

Swift and iOS development, SwiftUI architecture, async tests, SDK privacy, Xcode/SPM builds, and asset-catalog icons. Use for Swift/iOS tasks or test, build, icon actions. Not for web/Android assets or general raster editing. Output: target-compatible implementation or review, verification evidence, or an imageset.

13 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use cpp.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • testdino-hqSkill

    cpp

    Google's official C++ style guide. Covers headers, naming conventions, formatting, classes, memory management, RAII, smart pointers, and modern C++ features.

    Also in 4 projects that use this

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 0 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 11 edited versions we read keep these.

  • 0

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Use $swift <task>; optional test, build, and icon actions select the same internal task paths.

    line 12
  2. Very specific:

    Read local references from this skill directory.

    line 44
  3. Very specific:

    Run the project's test-first workflow for business behavior.

    line 52

“Common” is measured: kept by at least 6 in 10 of the 11 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Swift

Use $swift <task>; optional test, build, and icon actions select the same internal task paths. Complete the requested task with the smallest local reference set; specialist references are procedures, not independent skills.

Resolve the target before loading

  1. Establish the target package, Xcode project/scheme, or asset catalog from the request and project evidence. Ask and wait when a monorepo has several plausible targets. An intentional explicit target/version override wins; report the override separately from the observed configuration.
  2. For version-sensitive work read the target's SWIFT_VERSION, SWIFT_STRICT_CONCURRENCY, deployment target, and applicable Package.swift tools version. The installed swift --version is execution-environment evidence, not target compatibility. Preserve the precision of the evidence.
  3. Missing, conflicting, or broad incompatible constraints require a question and an answer before version-specific references or dependent actions. A migration needs an explicit destination version. For several destinations, group resources, changes, and verification by target.

Select only the task's resources

TaskLocal procedureLoad additional references when needed
Swift language/concurrency/errorslanguageconcurrency, value/reference, error handling; approachable-concurrency only for resolved Swift 6.2+
SwiftUI state/navigation/interopUIobservation-state, navigation, mvvm-coordinator, interop
Test authoring and strategy (test)testingswift-testing-api, xctest-bridging, xcuitest, snapshot-testing, test-taxonomy, protocol-di-host-testing
iOS SDK persistence/privacy/crypto/OCRplatformmatching SDK reference only
Xcode/SPM/lint/CI (build)buildswiftlint-config, xcodebuild-spm, scheme-signing
In-app symbol imageset (icon)iconlocal icon scripts below

Language rules also apply to code produced by another path; load the relevant language reference only if its compatibility or concurrency rules are needed. Read local references from this skill directory. Project source/configuration and external tools are explicit task inputs; a missing bundled resource is BLOCKED, not a reason to search an installed peer or repository module.

Actions and authority

Implementation requests authorize scoped source/tests/configuration edits; review and build/check requests stay read-only apart from explicit build outputs. Run the project's test-first workflow for business behavior. Retain separate local build, simulator/device, SDK/provider, and visual evidence. Missing tools produce NOT_RUN or UNAVAILABLE, never a successful verification.

For icons, agree on the requested catalog, name, source, color, and dimensions; search and preview before generating. Run scripts/iconify_gen.sh for Iconify (network + curl + a rasterizer) or scripts/generate_icons.swift for SF Symbols (macOS). Resolve paths from this skill directory and pass the explicit consumer catalog using --output. Verify 1x/2x/3x PNGs and Contents.json, then inspect the 2x PNG visually. These scripts generate in-app .imageset glyphs; an AppIcon request needs its own size matrix. Generation grants no commit or publication permission. Preserve external capability failures as non-pass evidence.

Completion

Report the resolved target/version and selected specialty; the resulting code, review findings, build commands/configuration, three-layer test plan, or imageset path; checks actually run and their verdicts; and unresolved compatibility, privacy, device, or visual gaps. Each specialist procedure retains its output and verification contract. A focused test result is separate from device or Host acceptance.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.