Sensorkit logo

Sensorkit

CommunityPopular
dpearson2699
sensorkit

Access research-grade sensor data using SensorKit for approved studies. Use when an app needs SensorKit entitlement setup, Research Sensor & Usage Data authorization, ambient light, recorded motion, device usage, keyboard metrics, visits, speech, face, wrist temperature, ECG, PPG, acoustic settings, or sleep-session data. Route ordinary motion to CoreMotion and health records/workouts to HealthKit.

Overview

Publisherdpearson2699
Repositoryswift-ios-skills
Skill namesensorkit
Stars
1.2K
Forks
61
Bundled files
2
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 2 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by dpearson2699 on GitHub. Read the source before you install it.

Installation

Install the Sensorkit AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/dpearson2699/swift-ios-skills/tree/main/skills/sensorkit
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/dpearson2699/swift-ios-skills.git /tmp/swift-ios-skills
mkdir -p .claude/skills
cp -r /tmp/swift-ios-skills/skills/sensorkit .claude/skills/sensorkit
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Sensorkit in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Sensorkit on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Sensorkit is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

SensorKit

Choose the exact SRSensor and verify its individual availability. Use CoreMotion for ordinary motion/activity features and HealthKit for health records and workouts.

Contents

Overview and Requirements

SensorKit enables research apps to record and fetch sensor data across iPhone and Apple Watch. The framework requires:

  1. Apple-approved research study -- submit a proposal at researchandcare.org.
  2. SensorKit entitlement -- Apple grants com.apple.developer.sensorkit.reader.allow only for approved studies.
  3. Manual provisioning profile -- Xcode requires an explicit App ID with the SensorKit capability enabled.
  4. User authorization -- the system presents a Research Sensor & Usage Data sheet that users approve per-sensor.
  5. Delayed retrieval -- design fetch timing around the canonical Data Holding Period.

An app can access up to 7 days of prior recorded data for an active sensor.

Entitlements

Add the SensorKit reader entitlement to a .entitlements file. List only the sensors Apple approved for the study. Common entitlement values include:

xml
<key>com.apple.developer.sensorkit.reader.allow</key>
<array>
    <string>ambient-light-sensor</string>
    <string>motion-accelerometer</string>
    <string>device-usage</string>
    <string>keyboard-metrics</string>
</array>

Load the Entitlement and Usage-Detail Catalog when selecting the exact entitlement string and NSSensorKitUsageDetail key for each approved sensor. Recheck specialized sensors against their individual SRSensor pages.

For manual signing, set Code Signing Entitlements to the entitlements file, Code Signing Identity to Apple Developer, Code Signing Style to Manual, and Provisioning Profile to the explicit profile with SensorKit capability.

Info.plist Configuration

Three keys are required:

xml
<!-- Study purpose shown in the authorization sheet -->
<key>NSSensorKitUsageDescription</key>
<string>This study monitors activity patterns for sleep research.</string>

<!-- Link to your study's privacy policy -->
<key>NSSensorKitPrivacyPolicyURL</key>
<string>https://example.com/privacy-policy</string>

<!-- Per-sensor usage explanations -->
<key>NSSensorKitUsageDetail</key>
<dict>
    <key>SRSensorUsageMotion</key>
    <dict>
        <key>Description</key>
        <string>Measures physical activity levels during the study.</string>
        <key>Required</key>
        <true/>
    </dict>
    <key>SRSensorUsageAmbientLightSensor</key>
    <dict>
        <key>Description</key>
        <string>Records ambient light to assess sleep environment.</string>
    </dict>
</dict>

If Required is true and the user denies that sensor, the system warns them that the study needs it and offers a chance to reconsider.

Use the exact usage-detail dictionary for each requested sensor. Load the Entitlement and Usage-Detail Catalog when mapping sensors beyond the motion and ambient-light examples above.

Authorization

Request authorization for the sensors your study needs. The system shows the Research Sensor & Usage Data sheet on first request.

swift
import SensorKit

let reader = SRSensorReader(sensor: .ambientLightSensor)

// Request authorization for multiple sensors at once
SRSensorReader.requestAuthorization(
    sensors: [.ambientLightSensor, .accelerometer, .keyboardMetrics]
) { error in
    if let error {
        print("Authorization request failed: \(error)")
    }
}

Use one status handler both for the initial check and delegate changes:

swift
private func applyAuthorizationStatus(
    _ status: SRAuthorizationStatus,
    to reader: SRSensorReader
) {
    switch status {
    case .authorized:
        reader.startRecording()
    case .denied:
        reader.stopRecording()
        // Direct the user to Settings > Privacy > Research Sensor & Usage Data.
    case .notDetermined:
        break // Request authorization first.
    @unknown default:
        break
    }
}

applyAuthorizationStatus(reader.authorizationStatus, to: reader)

func sensorReader(_ reader: SRSensorReader, didChange authorizationStatus: SRAuthorizationStatus) {
    applyAuthorizationStatus(authorizationStatus, to: reader)
}

Available Sensors

Load the Sensor Catalog to map each SRSensor to its sample type. Request only sensors approved for the study and recheck the selected sensor's availability and usage-detail key.

SRSensorReader

SRSensorReader is the central class for accessing sensor data. Each instance reads from a single sensor.

swift
import SensorKit

// Create a reader for one sensor
let lightReader = SRSensorReader(sensor: .ambientLightSensor)
let keyboardReader = SRSensorReader(sensor: .keyboardMetrics)

// Assign delegate to receive callbacks
lightReader.delegate = self
keyboardReader.delegate = self

The reader communicates through SRSensorReaderDelegate. Load the Delegate Method Catalog when wiring the complete authorization, recording, device-fetch, and sample-fetch lifecycle.

Recording and Fetching Data

Start and Stop Recording

swift
// Begin recording -- sensor stays active as long as any app has a stake
reader.startRecording()

// Stop recording -- framework deactivates the sensor when
// no app or system process is using it
reader.stopRecording()

Fetch Data

Build an SRFetchRequest with a time range and target device, then pass it to the reader:

swift
let request = SRFetchRequest()
request.device = SRDevice.current
request.from = SRAbsoluteTime(CFAbsoluteTimeGetCurrent() - 86400 * 2)  // 2 days ago
request.to = SRAbsoluteTime.current()

reader.fetch(request)

Receive results through the delegate:

swift
func sensorReader(
    _ reader: SRSensorReader,
    fetching request: SRFetchRequest,
    didFetchResult result: SRFetchResult<AnyObject>
) -> Bool {
    let timestamp = result.timestamp

    switch reader.sensor {
    case .ambientLightSensor:
        if let sample = result.sample as? SRAmbientLightSample {
            let lux = sample.lux
            let chromaticity = sample.chromaticity
            let placement = sample.placement
            processSample(lux: lux, chromaticity: chromaticity, at: timestamp)
        }
    case .keyboardMetrics:
        if let sample = result.sample as? SRKeyboardMetrics {
            let words = sample.totalWords
            let speed = sample.typingSpeed
            processKeyboard(words: words, speed: speed, at: timestamp)
        }
    case .deviceUsageReport:
        if let sample = result.sample as? SRDeviceUsageReport {
            let wakes = sample.totalScreenWakes
            let unlocks = sample.totalUnlocks
            processUsage(wakes: wakes, unlocks: unlocks, at: timestamp)
        }
    default:
        break
    }

    return true  // Return true to continue receiving results
}

func sensorReader(_ reader: SRSensorReader, didCompleteFetch request: SRFetchRequest) {
    print("Fetch complete for \(reader.sensor)")
}

func sensorReader(
    _ reader: SRSensorReader,
    fetching request: SRFetchRequest,
    failedWithError error: any Error
) {
    print("Fetch failed: \(error)")
}

Cast result.sample to the sample shape for the reader's sensor. Some streams return one object per result, while recorded motion, ECG, PPG, and ambient pressure streams can return arrays of recorded samples.

Data Holding Period

SensorKit imposes a 24-hour holding period on newly recorded data. Fetch requests whose time range overlaps this period return no results. Design data collection workflows around this delay.

SRDevice

SRDevice identifies the hardware source for sensor samples. Use it to distinguish data from iPhone versus Apple Watch.

swift
// Get the current device
let currentDevice = SRDevice.current
print("Model: \(currentDevice.model)")
print("System: \(currentDevice.systemName) \(currentDevice.systemVersion)")

// Fetch all available devices for a sensor
reader.fetchDevices()

Handle fetched devices through the delegate:

swift
func sensorReader(_ reader: SRSensorReader, didFetch devices: [SRDevice]) {
    for device in devices {
        let request = SRFetchRequest()
        request.device = device
        request.from = SRAbsoluteTime(CFAbsoluteTimeGetCurrent() - 86400)
        request.to = SRAbsoluteTime.current()
        reader.fetch(request)
    }
}

func sensorReader(_ reader: SRSensorReader, fetchDevicesDidFailWithError error: any Error) {
    print("Failed to fetch devices: \(error)")
}

SRDevice Properties

PropertyTypeDescription
modelStringUser-defined device name
nameStringFramework-defined device name
systemNameStringOS name (iOS, watchOS)
systemVersionStringOS version
productTypeStringHardware identifier
currentSRDeviceClass property for the running device

Common Mistakes

DON'T: Attempt to use SensorKit without the entitlement

Obtain Apple's study approval, the sensor-specific entitlement values, and a matching manual provisioning profile before constructing production readers.

DON'T: Expect immediate data access

Apply the Data Holding Period; a fetch overlapping the hold is not proof that recording failed.

DON'T: Forget to set the delegate before fetching

Assign the delegate before startRecording() or fetch(_:); results and failures arrive through delegate callbacks.

DON'T: Skip per-sensor Info.plist usage detail

Add the exact Info.plist Configuration usage-detail entry for every requested sensor.

DON'T: Ignore SRError codes

Distinguish at least .invalidEntitlement, .noAuthorization, .dataInaccessible, .fetchRequestInvalid, .promptDeclined, and unknown future codes. Load the Full Delegate Implementation for the complete switch and callback wiring.

Review Checklist

  • Apple-approved research study in place before development
  • com.apple.developer.sensorkit.reader.allow entitlement lists only needed sensors
  • Manual provisioning profile with explicit App ID and SensorKit capability
  • NSSensorKitUsageDescription in Info.plist with clear study purpose
  • NSSensorKitPrivacyPolicyURL in Info.plist with valid privacy policy URL
  • NSSensorKitUsageDetail entries for every requested sensor
  • Required key set appropriately for essential vs. optional sensors
  • Authorization requested before recording, status checked before fetching
  • Delegate assigned before calling startRecording() or fetch(_:)
  • Fetch request time ranges account for 24-hour data holding period
  • SRError codes handled in all failure delegate methods
  • fetchDevices() used to discover available devices before fetching
  • stopRecording() called when data collection is complete
  • sensorReader(_:fetching:didFetchResult:) returns true to continue or false to stop

References

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Sensorkit AI skill do?

Access research-grade sensor data using SensorKit for approved studies. Use when an app needs SensorKit entitlement setup, Research Sensor & Usage Data authorization, ambient light, recorded motion, device usage, keyboard metrics, visits, speech, face, wrist temperature, ECG, PPG, acoustic settings, or sleep-session data. Route ordinary motion to CoreMotion and health records/workouts to HealthKit.

Why use Sensorkit on TypingMind?

Because you install it once and use it with any model. Sensorkit is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Sensorkit in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/dpearson2699/swift-ios-skills/tree/main/skills/sensorkit. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Sensorkit?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Sensorkit?

As many as you like. As long as a model supports skills, you can use Sensorkit with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Sensorkit AI skill free?

It is published on GitHub by dpearson2699. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇