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.
npx skills add troyjthomas/ios-agent-skills --skill mcp-setupgit clone --depth 1 https://github.com/troyjthomas/ios-agent-skillsWrote 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/skills/troyjthomas/ios-agent-skills/mcp-setup)<a href="https://agentmods.dev/skills/troyjthomas/ios-agent-skills/mcp-setup"><img src="https://agentmods.dev/badge/skills/troyjthomas/ios-agent-skills/mcp-setup/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/troyjthomas/ios-agent-skills/mcp-setup"><img src="https://agentmods.dev/badge/skills/troyjthomas/ios-agent-skills/mcp-setup.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00091 | $0.01386 |
| Opus 5.5 | $0.00036 | $0.00554 |
| Sonnet 5.5 | $0.00018 | $0.00277 |
| Haiku 4.5 | $0.00009 | $0.00139 |
Grade A, and why
mcp-setup scanned grade A with 1 finding 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 1mo 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -o ~/.claude/skills/sosumi/SKILL.md https://sosumi.ai/SKILL.md How it starts
The opening of the file, as written. The whole thing — 141 lines — stays where its author put it; the contents beside it link to each section on GitHub.
MCP Setup
Four MCPs give Claude Code full access to your design files, Apple documentation, the Xcode build system, and visual preview rendering. Set these up once and every project benefits.
The Stack
| MCP | What It Does | When Claude Uses It |
|---|---|---|
| Figma MCP | Reads your Figma designs directly | Building/matching designed screens |
| Sosumi | Apple docs, HIG, WWDC sessions as markdown | Looking up APIs, guidelines, Liquid Glass patterns |
| XcodeBuildMCP | Build, test, simulator, deploy (59 tools) | Building the project, running tests, deploying |
| Xcode MCP Bridge | SwiftUI previews, diagnostics, doc search | Visually verifying what it built, fixing errors |
Setup Instructions
1. Figma MCP
You likely already have this connected. If not:
claude mcp add figma -- npx -y figma-developer-mcp --figma-api-key YOUR_KEY
Or connect via OAuth in Claude Code settings. This lets Claude Code read frame layouts, text content, colors, spacing, and component hierarchy directly from your Figma files.
2. Sosumi (Apple Developer Documentation)
This is critical. Apple's docs are locked behind JavaScript, which means Claude Code can't read them by default. Sosumi translates Apple Developer docs, Human Interface Guidelines, and WWDC sessions into AI-friendly Markdown.
claude mcp add sosumi -- npx -y mcp-remote https://sosumi.ai/mcp
Or as a Claude Code skill (available in every project):
mkdir -p ~/.claude/skills/sosumi
curl -o ~/.claude/skills/sosumi/SKILL.md https://sosumi.ai/SKILL.md
What this enables:
- Claude Code can look up any SwiftUI modifier, any HIG guideline
- When building Liquid Glass effects, it reads the actual iOS 26 docs
- When you say "use the correct presentation style," it checks HIG
- WWDC session transcripts are searchable for implementation guidance
3. XcodeBuildMCP (Build System)
This gives Claude Code 59 tools for building, testing, and deploying without Xcode open.
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.
- 1mo ago First seen · 141 lines · 91 tokens per session scan A 504afc3a80b6
mcp-setup is a skill published in the GitHub repository troyjthomas/ios-agent-skills (2 stars, last pushed 5mo ago), licensed MIT. It adds 91 tokens to every session and 1,386 once invoked, about $0.0004 per session on Opus 5.5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
react-native-gen
React Native code generation from Memoire specs — NativeWind styling, Expo-compatible, platform-specific patterns.
swiftui-design-engineering
Design, build, refactor, and review production SwiftUI interfaces with explicit state ownership, navigation, accessibility, design tokens, previews, and simulator verification.
swift-concurrency-safety
Design and review Swift 6 concurrency boundaries using structured tasks, actors, Sendable values, cancellation, and compiler-backed data-race safety.
swiftui-liquid-glass
Adopt or audit iOS 26+ SwiftUI Liquid Glass with native APIs, correct availability fallbacks, restrained hierarchy, accessibility, and measured rendering performance.
xcode-build-reliability
Make Xcode builds agent-verifiable with deterministic scheme discovery, concise logs, explicit destinations, stable scripts, build-time evidence, and CI parity.
ios-app-intents
Design, implement, and verify App Intents, App Entities, App Shortcuts, widgets, Siri, Spotlight, and Apple Intelligence surfaces with stable identity and safe execution.