Event Hooks

swift-agents

2 events · 2 hooks

npx claudepluginhub techopolis/swift-agents

Source

Defined in hooks/hooks.json

Configuration

SubagentStop

Hooks:

command

cat <<'HOOK_OUTPUT'
VALIDATION: Before presenting specialist output to user:
- Check for conflicts between specialist recommendations
- Verify accessibility was reviewed for UI code
- Reject any mock, placeholder, or stub implementations
HOOK_OUTPUT

UserPromptSubmit

Hooks:

command

cat <<'HOOK_OUTPUT'
INSTRUCTION: SWIFT AGENTS CHECK
If this task involves Swift, SwiftUI, UIKit, async/await, actors, Core ML,
MLX, Foundation Models, SwiftData, StoreKit, visionOS, RealityKit, Keychain,
CryptoKit, App Store submission, or any iOS/macOS/visionOS application code:

1. Delegate to the swift-lead agent.
2. The swift-lead determines which specialists review the task.
3. Do NOT write UI code without mobile-a11y-specialist review.
4. Do NOT write concurrent code without concurrency-specialist review.
5. NEVER create mock, placeholder, or stub implementations.
6. Use built-in SwiftUI features instead of custom replacements.

If the task does not involve Swift or Apple platform code, proceed normally.
HOOK_OUTPUT

Summary

Injects Swift/iOS delegation instructions on UserPromptSubmit and validation checks (conflicts, UI accessibility, no mocks/stubs) on SubagentStop for Apple platform workflows.