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 agentmods add skills/junmystery/agent-guidance-python/foundation-models-on-devicenpx skills add JunMystery/Agent-Guidance-Python --skill foundation-models-on-devicegit clone --depth 1 https://github.com/JunMystery/Agent-Guidance-PythonWrote 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/junmystery/agent-guidance-python/foundation-models-on-device)<a href="https://agentmods.dev/skills/junmystery/agent-guidance-python/foundation-models-on-device"><img src="https://agentmods.dev/badge/skills/junmystery/agent-guidance-python/foundation-models-on-device.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 | $0.00038 | $0.01731 |
| Opus 5 | $0.00019 | $0.00865 |
| Sonnet 5 | $0.00008 | $0.00346 |
| Haiku 4.5 | $0.00004 | $0.00173 |
Grade A, and why
foundation-models-on-device 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 2d 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.
This is a copy
97% identical to foundation-models-on-device — 23 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 244 lines — stays where its author put it; the contents beside it link to each section on GitHub.
FoundationModels: On-Device LLM (iOS 26)
Patterns for integrating Apple's on-device language model into apps using the FoundationModels framework. Covers text generation, structured output with @Generable, custom tool calling, and snapshot streaming — all running on-device for privacy and offline support.
When to Activate
- Building AI-powered features using Apple Intelligence on-device
- Generating or summarizing text without cloud dependency
- Extracting structured data from natural language input
- Implementing custom tool calling for domain-specific AI actions
- Streaming structured responses for real-time UI updates
- Need privacy-preserving AI (no data leaves the device)
Core Pattern — Availability Check
Always check model availability before creating a session:
struct GenerativeView: View {
private var model = SystemLanguageModel.default
var body: some View {
switch model.availability {
case .available:
ContentView()
case .unavailable(.deviceNotEligible):
Text("Device not eligible for Apple Intelligence")
case .unavailable(.appleIntelligenceNotEnabled):
Text("Please enable Apple Intelligence in Settings")
case .unavailable(.modelNotReady):
Text("Model is downloading or not ready")
case .unavailable(let other):
Text("Model unavailable: \(other)")
}
}
}
Core Pattern — Basic Session
// Single-turn: create a new session each time
let session = LanguageModelSession()
let response = try await session.respond(to: "What's a good month to visit Paris?")
print(response.content)
// Multi-turn: reuse session for conversation context
let session = LanguageModelSession(instructions: """
You are a cooking assistant.
Provide recipe suggestions based on ingredients.
Keep suggestions brief and practical.
""")
let first = try await session.respond(to: "I have chicken and rice")
let followUp = try await session.respond(to: "What about a vegetarian option?")
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.
- 2d ago First seen · 244 lines · 38 tokens per session scan A 8efee7e392b3
foundation-models-on-device is a skill published in the GitHub repository JunMystery/Agent-Guidance-Python (2 stars, last pushed 1mo ago), licensed MIT. It adds 38 tokens to every session and 1,731 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. It is 97% identical to foundation-models-on-device, differing in 23 lines, and is treated as a copy.
Other skills, from other repositories
debug-evals
Debug haiku.rag evaluation runs in Logfire. Use when asked to look at Logfire for an eval run, find failing or low-scoring eval cases, compare runs, check citation quality (citedmap) or judge pass rate (answerequivalent), or explain why an eval case failed. Drives the Logfire MCP against the evals service, or the…
android-navigation-3
Install and migrate to Jetpack Navigation 3. Use when implementing Navigation 3 patterns including NavDisplay, NavKey routes, deep links, multiple backstacks, scenes (dialogs, bottom sheets), or migrating from Navigation 2.
td-api-reference
MUST READ before writing TD Python via executepython, setdatcontent, or editdatcontent, and before any heavy build. Parameters, storage, operators, referencing, threading, cook model, heavy-build safety caps.
aba-precision-protocol
FOUNDATIONAL BEHAVIORAL PROTOCOL — ABA-based precision execution rules. Every prompt processed must follow these rules. Mistakes caught late create intermittent reinforcement of wrong patterns. Stop-fix-verify before proceeding.
android-compose-migration
Migrate an Android XML View to Jetpack Compose following a structured 10-step workflow. Use when converting XML layouts to Compose, setting up Compose in an existing View-based project, or incrementally adopting Compose.
common-mobile-animation
Apply motion design principles for mobile apps covering timing curves, transitions, gestures, and performance-conscious animations. Use when implementing screen transitions, gesture-driven interactions, shared-element animations, or optimizing animation frame rates on iOS, Android, or Flutter.