Skip to content

Latest commit

 

History

History
145 lines (104 loc) · 7.23 KB

File metadata and controls

145 lines (104 loc) · 7.23 KB

AGENT.md

Behavioral guidelines to reduce common LLM coding mistakes. Merge with project-specific instructions as needed.

Tradeoff: These guidelines bias toward caution over speed. For trivial tasks, use judgment.

1. Think Before Coding

Don't assume. Don't hide confusion. Surface tradeoffs.

Before implementing:

  • State your assumptions explicitly. If uncertain, ask.
  • If multiple interpretations exist, present them - don't pick silently.
  • If a simpler approach exists, say so. Push back when warranted.
  • If something is unclear, stop. Name what's confusing. Ask.

2. Simplicity First

Minimum code that solves the problem. Nothing speculative.

  • No features beyond what was asked.
  • No abstractions for single-use code.
  • No "flexibility" or "configurability" that wasn't requested.
  • No error handling for impossible scenarios.
  • If you write 200 lines and it could be 50, rewrite it.

Ask yourself: "Would a senior engineer say this is overcomplicated?" If yes, simplify.

3. Surgical Changes

Touch only what you must. Clean up only your own mess.

When editing existing code:

  • Don't "improve" adjacent code, comments, or formatting.
  • Don't refactor things that aren't broken.
  • Match existing style, even if you'd do it differently.
  • If you notice unrelated dead code, mention it - don't delete it.

When your changes create orphans:

  • Remove imports/variables/functions that YOUR changes made unused.
  • Don't remove pre-existing dead code unless asked.

The test: Every changed line should trace directly to the user's request.

4. Goal-Driven Execution

Define success criteria. Loop until verified.

Transform tasks into verifiable goals:

  • "Add validation" → "Write tests for invalid inputs, then make them pass"
  • "Fix the bug" → "Write a test that reproduces it, then make it pass"
  • "Refactor X" → "Ensure tests pass before and after"

For multi-step tasks, state a brief plan:

1. [Step] → verify: [check]
2. [Step] → verify: [check]
3. [Step] → verify: [check]

Strong success criteria let you loop independently. Weak criteria ("make it work") require constant clarification.


These guidelines are working if: fewer unnecessary changes in diffs, fewer rewrites due to overcomplication, and clarifying questions come before implementation rather than after mistakes.

Core instructions

  • Target macOS 15.2 or later.
  • Swift 6.2 or later, using modern Swift concurrency.
  • SwiftUI backed up by @Observable classes for shared data.
  • Do not introduce third-party frameworks without asking first.
  • Avoid UIKit unless requested.

Swift instructions

  • Always mark @Observable classes with @MainActor.
  • Assume strict Swift concurrency rules are being applied.
  • Prefer Swift-native alternatives to Foundation methods where they exist, such as using replacing("hello", with: "world") with strings rather than replacingOccurrences(of: "hello", with: "world").
  • Prefer modern Foundation API, for example URL.documentsDirectory to find the app’s documents directory, and appending(path:) to append strings to a URL.
  • Never use C-style number formatting such as Text(String(format: "%.2f", abs(myNumber))); always use Text(abs(change), format: .number.precision(.fractionLength(2))) instead.
  • Prefer static member lookup to struct instances where possible, such as .circle rather than Circle(), and .borderedProminent rather than BorderedProminentButtonStyle().
  • Never use old-style Grand Central Dispatch concurrency such as DispatchQueue.main.async(). If behavior like this is needed, always use modern Swift concurrency.
  • Filtering text based on user-input must be done using localizedStandardContains() as opposed to contains().
  • Avoid force unwraps and force try unless it is unrecoverable.

SwiftUI instructions

  • Always use foregroundStyle() instead of foregroundColor().
  • Always use clipShape(.rect(cornerRadius:)) instead of cornerRadius().
  • Always use the Tab API instead of tabItem().
  • Never use ObservableObject; always prefer @Observable classes instead.
  • Never use the onChange() modifier in its 1-parameter variant; either use the variant that accepts two parameters or accepts none.
  • Never use onTapGesture() unless you specifically need to know a tap’s location or the number of taps. All other usages should use Button.
  • Never use Task.sleep(nanoseconds:); always use Task.sleep(for:) instead.
  • Never use UIScreen.main.bounds to read the size of the available space.
  • Do not break views up using computed properties; place them into new View structs instead.
  • Do not force specific font sizes; prefer using Dynamic Type instead.
  • Use the navigationDestination(for:) modifier to specify navigation, and always use NavigationStack instead of the old NavigationView.
  • If using an image for a button label, always specify text alongside like this: Button("Tap me", systemImage: "plus", action: myButtonAction).
  • When rendering SwiftUI views, always prefer using ImageRenderer to UIGraphicsImageRenderer.
  • Don’t apply the fontWeight() modifier unless there is good reason. If you want to make some text bold, always use bold() instead of fontWeight(.bold).
  • Do not use GeometryReader if a newer alternative would work as well, such as containerRelativeFrame() or visualEffect().
  • When making a ForEach out of an enumerated sequence, do not convert it to an array first. So, prefer ForEach(x.enumerated(), id: \.element.id) instead of ForEach(Array(x.enumerated()), id: \.element.id).
  • When hiding scroll view indicators, use the .scrollIndicators(.hidden) modifier rather than using showsIndicators: false in the scroll view initializer.
  • Place view logic into view models or similar, so it can be tested.
  • Avoid AnyView unless it is absolutely required.
  • Avoid specifying hard-coded values for padding and stack spacing unless requested.
  • Avoid using UIKit colors in SwiftUI code.

SwiftData instructions

If SwiftData is configured to use CloudKit:

  • Never use @Attribute(.unique).
  • Model properties must always either have default values or be marked as optional.
  • All relationships must be marked optional.

Project structure

  • Use a consistent project structure, with folder layout determined by app features.
  • Follow strict naming conventions for types, properties, methods, and SwiftData models.
  • Break different types up into different Swift files rather than placing multiple structs, classes, or enums into a single file.
  • Write unit tests for core application logic.
  • Only write UI tests if unit tests are not possible.
  • Add code comments and documentation comments as needed.
  • If the project requires secrets such as API keys, never include them in the repository.

Git workflow

  • When working on a GitHub issue, use gh issue develop <issue-number> --checkout to create and checkout a branch with consistent naming.
  • Use conventional commit format for all commits: type(scope): subject
    • Types: feat, fix, docs, style, refactor, test, chore
    • Example: feat(menu-bar): add open output folder button
    • Keep subject line under 50 characters when possible
    • Use present tense ("add" not "added")

PR instructions

  • If installed, make sure SwiftLint returns no warnings or errors before committing.