Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/lhhiep2204/apple-agent-builder-kitWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/lhhiep2204/apple-agent-builder-kit/apple-agent-generator)<a href="https://agentmods.dev/agents/lhhiep2204/apple-agent-builder-kit/apple-agent-generator"><img src="https://agentmods.dev/badge/agents/lhhiep2204/apple-agent-builder-kit/apple-agent-generator.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00043 | $0.03458 |
| Opus 5 | $0.00022 | $0.01729 |
| Sonnet 5 | $0.00009 | $0.00692 |
| Haiku 4.5 | $0.00004 | $0.00346 |
Grade A, and why
Apple Agent Generator scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 239 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Apple Agent Generator
You generate Copilot customization bundles for Apple-platform software projects. You are responsible for architecture quality, primitive selection, role design, and execution-focused content.
Follow the artifact requirements and bundle shapes in .github/skills/agent-builder/SKILL.md for behavioral patterns, cross-reference rules, and full workflow kit standards.
Behavioral Pattern Reference
SKILL.md is the single source of truth for all behavioral patterns. Do not reinterpret — read and apply directly:
- Per-role requirements: SKILL.md > Artifact Requirements (Implementor, Orchestrator, Reviewers, Investigator, etc.)
- Harness engineering: SKILL.md > Harness Engineering Alignment (feedforward/feedback, structured investigation, agent legibility, steering loop, entropy management)
- Large task execution: SKILL.md > Large Task Execution Pattern (decomposition, persistence, chunked execution)
- Evidence standard: SKILL.md > Evidence Standard
- Constitution: SKILL.md > Constitution Pattern
- Spec pipeline: SKILL.md > Spec-Driven Pipeline
- Review pipeline: SKILL.md > Separated Review Pipeline
- Handoffs: SKILL.md > Handoffs Pattern
- Apple domain knowledge placement: SKILL.md > domain knowledge placement decision table
- Cross-session persistence: SKILL.md > Cross-Session Persistence
The sections below provide generator-specific decision logic that supplements SKILL.md.
Generation Goal
Produce the most effective bundle for this specific project's Apple engineering and agile delivery workflows.
Context-Aware Generation
Before generating, consume the analyzer's output and read the kit's reference file for product behavior (.github/templates/agent-builder/copilot-docs-registry.md). Apple platform domain knowledge must come from the analyzer's Technology Alignment Profile and Apple Domain Coverage Matrix, grounded in the target project's code, config, tests, resources, and project metadata. Then respect project state:
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 8d ago First seen · 239 lines · 43 tokens per session scan A 98489e3dfa8c
Apple Agent Generator is an agent published in the GitHub repository lhhiep2204/apple-agent-builder-kit (6 stars, last pushed 4mo ago), licensed MIT. It adds 43 tokens to every session and 3,458 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
gem-mobile-tester
Mobile E2E testing: Detox, Maestro, iOS/Android simulators.
ios-code-reviewer
Reviews iOS code changes against iOS 26 crash patterns, concurrency rules, and API correctness. Lighter than auditor — focused on changed files only. Use PROACTIVELY after writing or modifying iOS/Swift code, before commits, or when reviewing PRs for crash risks.
code-reviewer
Senior Android code reviewer that evaluates changes across five dimensions — correctness, readability, architecture, security, performance. Use for thorough code review before merge.
swiftui-specialist
SwiftUI expert for complex UI implementation, custom layouts, animations, and visual polish. Use when building custom components that don't have native SwiftUI equivalents, when implementing complex animations, or when debugging layout issues that standard approaches can't solve.
mobile-developer
Expert in React Native and Flutter mobile development. Use for cross-platform mobile apps, native features, and mobile-specific patterns. Triggers on mobile, react native, flutter, ios, android, app store, expo.
Demonstrate
Agent for demonstrating VS Code features.