[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7F174DR3F9MPSP63C0G124P":3,"skill-same:01M7F174DR3F9MPSP63C0G124P":54},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":26,"commitSha":27,"skillDirectory":28,"forksCount":14,"repositoryHtmlUrl":29,"repositoryHomepageUrl":6,"ownerHtmlUrl":30,"ownerBlogUrl":31,"ownerDescription":32,"updatedAt":33,"install":34,"resources":47},"01M7F174DR3F9MPSP63C0G124P","ios-on-device-intelligence",null,"Design, build, evaluate, or review Apple on-device intelligence features — Foundation Models, Vision, Core ML, Speech, Translation, Natural Language, Sound Analysis, and App Intents — with explicit availability, typed and validated output, privacy boundaries, review before side effects, and device evaluation. Use when a feature generates, extracts, classifies, transcribes, translates, or acts on user content with an Apple model or framework.","shotcowboystyle\u002Fag-skills","shotcowboystyle","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F62471?v=4","01M7EZTCFCZJ6K3MYW8BPM4X6R","Curtis Blanton","01M7EZTCG1S5M04J90A931X9BX",0,"MIT","https:\u002F\u002Fgithub.com\u002Fshotcowboystyle\u002Fag-skills\u002Fblob\u002Fmain\u002Fplugins\u002Fios-ops\u002Fskills\u002Fios-on-device-intelligence\u002FSKILL.md",[],[],"2026-11-16 10:05:34",false,"---\nname: ios-on-device-intelligence\ndescription: Design, build, evaluate, or review Apple on-device intelligence features — Foundation Models, Vision, Core ML, Speech, Translation, Natural Language, Sound Analysis, and App Intents — with explicit availability, typed and validated output, privacy boundaries, review before side effects, and device evaluation. Use when a feature generates, extracts, classifies, transcribes, translates, or acts on user content with an Apple model or framework.\nallowed-tools: Read, Grep, Glob, Edit, Write, Bash\nlast_verified: 2026-08-19\n---\n\n# iOS On-Device Intelligence\n\n## Overview\n\nUse this skill to choose the narrowest intelligence route and keep availability, privacy, model output, deterministic validation, review, fallback, evaluation, and physical-device evidence explicit.\n\n`authorized source -> bounded input -> available model\u002Fasset -> cancellable proposal -> validation\u002Fprovenance -> review\u002Fauthorization -> durable result or side effect`\n\n## Read before acting\n\n- Inspect the actual Xcode targets, deployment target, device family, model resources, language assets, entitlements, usage descriptions, persistence, network\u002Fserver routes, and current AI adapter.\n- Read the [AI route selector](..\u002Fios-knowledge-base\u002Freferences\u002F30-on-device-ai\u002F00-ai-route-selector.md), [Foundation Models mental model](..\u002Fios-knowledge-base\u002Freferences\u002F30-on-device-ai\u002F01-foundation-models-mental-model.md), [privacy\u002Favailability\u002Ffallback guidance](..\u002Fios-knowledge-base\u002Freferences\u002F30-on-device-ai\u002F06-privacy-availability-and-fallback.md), [availability\u002Fproof matrix](..\u002Fios-knowledge-base\u002Freferences\u002F30-on-device-ai\u002F08-on-device-ai-availability-and-proof-matrix.md), and [evaluation\u002Fsafety\u002Ffallback recipe](..\u002Fios-knowledge-base\u002Freferences\u002F31-on-device-ai-recipes\u002F05-evaluation-safety-and-fallback.md).\n- This skill owns the model, its availability, and its output. When live capture, media import, or audio sessions feed the model, pair it with [ios-media-and-inputs](..\u002Fios-media-and-inputs\u002FSKILL.md) for the capture lifecycle.\n- Refresh the official Apple pages listed in `references\u002Fdetails.md` (last checked August 2026) before relying on model availability, device\u002Fregion\u002Flanguage behavior, API spelling, output safety, tool calling, context limits, or privacy claims.\n\n## Route workflow\n\n1. State the user outcome and decide whether a model is necessary. Use deterministic code for calculations, parsing, validation, routing, permissions, sorting, and safety rules; use a model only when ambiguity or learned perception\u002Flanguage is the value.\n2. Choose the narrowest route: Foundation Models for bounded language generation\u002Fextraction; Vision for system observations; Core ML for a packaged custom model; Speech for the exact transcription route; Translation for supported language pairs; Natural Language for deterministic text analysis; Sound Analysis for bounded audio classification; App Intents for system-discoverable actions.\n3. Record the route matrix: target SDK, OS\u002Fdevice eligibility, Apple Intelligence\u002Fmodel state, model\u002Fresource version, locale\u002Flanguage asset, permission, input sensitivity\u002Fretention, server or Private Cloud Compute boundary, expected uncertainty, and manual\u002Fdeterministic fallback.\n4. Model availability as state, not a Boolean: available, not eligible, disabled, model\u002Fasset not ready, unsupported language, permission denied, context exceeded, guardrail refusal, cancelled, input failure, service error, and retry\u002Fmanual route.\n5. Keep trusted instructions separate from user or external content. Bound input size, duration, image\u002Faudio dimensions, context, tool arguments, and concurrency. Version prompts, schemas, model revisions, preprocessing, and evaluation fixtures.\n6. Prefer typed\u002Fguided output for structured proposals. Validate every field, enum, range, identifier, source reference, confidence\u002Fquality value, and authorization before it can affect domain state, export, messaging, payment, health data, or an irreversible action.\n7. Keep tools narrow and app-owned. Prefer read-only tools; for mutation require deterministic authorization, validation, idempotence, timeout\u002Ferror handling, audit\u002Fprovenance, and explicit user confirmation. The model never becomes the authority for permissions or side effects.\n8. Store generated drafts, observations, predictions, transcripts, translations, and classifications separately from trusted domain truth. Show source context\u002Funcertainty when it matters, allow edit\u002Freject\u002Fretry, redact logs, and define raw-input\u002Foutput retention and deletion.\n9. Evaluate representative, empty, oversized, multilingual, adversarial, safety-sensitive, low-quality, and stale-input fixtures. Record quality, abstention\u002Frefusal, correction rate, latency, memory, battery, thermal state, dropped inputs, and device\u002FOS\u002Fmodel configuration.\n10. Verify the actual physical device and target language\u002Fmodel configuration for any claim that depends on Apple Intelligence availability, camera\u002Fmicrophone\u002Fsensor behavior, on-device performance, or system-surface invocation. A simulator, mock, preview, or successful compile is narrower evidence.\n\n## Fast path\n\nEstablish a deterministic baseline before tuning a model. Compare a representative fixture set covering valid output, malformed output, refusal, missing context, and oversized context, and promote only fields that pass schema, safety, and human-review gates. Expand the fixture set only when a failure reveals a new risk class.\n\n## Evaluation and safety contract\n\nFor each evaluation record:\n\n- route\u002Fframework\u002FAPI and source URL;\n- target SDK, OS\u002Fbuild, device family, region\u002Flanguage, Apple Intelligence\u002Fmodel\u002Fasset state, and model\u002Fprompt\u002Fschema version;\n- input source, redaction\u002Fminimization, maximum size, retention, and consent\u002Fpermission state;\n- expected behavior, fixture class, output type, quality\u002Fcorrection\u002Fabstention metric, latency, memory, battery, and thermal observations;\n- validation\u002Freview\u002Fcommit rule, fallback, and unresolved risks.\n\nRe-run representative prompts\u002Ffixtures after OS, SDK, model, language, prompt, schema, preprocessing, or tool changes. Test prompt-injection-like content when user or web content enters the context. Apple guardrails are a base layer; add app-specific domain, audience, privacy, and harm validation.\n\n## Non-negotiable safety and evidence rules\n\n- Never present generated text, OCR, classification, translation, transcription, sound analysis, or model output as medical\u002Flegal\u002Ffinancial truth, identity, authenticity, guaranteed accuracy, or authorization without domain validation and appropriate human review.\n- Never claim “on-device,” “private,” “real-time,” “supported,” “accurate,” or “safe” from an imported framework, model response, simulator, preview, mock, one prompt, one language, one device, or successful compile.\n- Keep source permission, model availability, input availability, output quality, domain validation, and side-effect authorization separate. A model that is ready can still return an unusable or unsafe proposal.\n- Do not send sensitive prompts, audio, images, health\u002Fpersonal data, or model output remotely or retain raw inputs beyond the stated feature. Do not add broad permissions, analytics, telemetry, accounts, or servers merely to make an AI demo work.\n- Cancel and clean up model sessions, capture, inference, translation, transcription, tool calls, and background tasks when the feature ends. Ignore stale results so an older proposal cannot overwrite newer user state.\n\nPer-framework boundaries, deliverable, change boundary, and sources: `references\u002Fdetails.md`.\n","# iOS On-Device Intelligence\n\n## Overview\n\nUse this skill to choose the narrowest intelligence route and keep availability, privacy, model output, deterministic validation, review, fallback, evaluation, and physical-device evidence explicit.\n\n`authorized source -> bounded input -> available model\u002Fasset -> cancellable proposal -> validation\u002Fprovenance -> review\u002Fauthorization -> durable result or side effect`\n\n## Read before acting\n\n- Inspect the actual Xcode targets, deployment target, device family, model resources, language assets, entitlements, usage descriptions, persistence, network\u002Fserver routes, and current AI adapter.\n- Read the [AI route selector](..\u002Fios-knowledge-base\u002Freferences\u002F30-on-device-ai\u002F00-ai-route-selector.md), [Foundation Models mental model](..\u002Fios-knowledge-base\u002Freferences\u002F30-on-device-ai\u002F01-foundation-models-mental-model.md), [privacy\u002Favailability\u002Ffallback guidance](..\u002Fios-knowledge-base\u002Freferences\u002F30-on-device-ai\u002F06-privacy-availability-and-fallback.md), [availability\u002Fproof matrix](..\u002Fios-knowledge-base\u002Freferences\u002F30-on-device-ai\u002F08-on-device-ai-availability-and-proof-matrix.md), and [evaluation\u002Fsafety\u002Ffallback recipe](..\u002Fios-knowledge-base\u002Freferences\u002F31-on-device-ai-recipes\u002F05-evaluation-safety-and-fallback.md).\n- This skill owns the model, its availability, and its output. When live capture, media import, or audio sessions feed the model, pair it with [ios-media-and-inputs](..\u002Fios-media-and-inputs\u002FSKILL.md) for the capture lifecycle.\n- Refresh the official Apple pages listed in `references\u002Fdetails.md` (last checked August 2026) before relying on model availability, device\u002Fregion\u002Flanguage behavior, API spelling, output safety, tool calling, context limits, or privacy claims.\n\n## Route workflow\n\n1. State the user outcome and decide whether a model is necessary. Use deterministic code for calculations, parsing, validation, routing, permissions, sorting, and safety rules; use a model only when ambiguity or learned perception\u002Flanguage is the value.\n2. Choose the narrowest route: Foundation Models for bounded language generation\u002Fextraction; Vision for system observations; Core ML for a packaged custom model; Speech for the exact transcription route; Translation for supported language pairs; Natural Language for deterministic text analysis; Sound Analysis for bounded audio classification; App Intents for system-discoverable actions.\n3. Record the route matrix: target SDK, OS\u002Fdevice eligibility, Apple Intelligence\u002Fmodel state, model\u002Fresource version, locale\u002Flanguage asset, permission, input sensitivity\u002Fretention, server or Private Cloud Compute boundary, expected uncertainty, and manual\u002Fdeterministic fallback.\n4. Model availability as state, not a Boolean: available, not eligible, disabled, model\u002Fasset not ready, unsupported language, permission denied, context exceeded, guardrail refusal, cancelled, input failure, service error, and retry\u002Fmanual route.\n5. Keep trusted instructions separate from user or external content. Bound input size, duration, image\u002Faudio dimensions, context, tool arguments, and concurrency. Version prompts, schemas, model revisions, preprocessing, and evaluation fixtures.\n6. Prefer typed\u002Fguided output for structured proposals. Validate every field, enum, range, identifier, source reference, confidence\u002Fquality value, and authorization before it can affect domain state, export, messaging, payment, health data, or an irreversible action.\n7. Keep tools narrow and app-owned. Prefer read-only tools; for mutation require deterministic authorization, validation, idempotence, timeout\u002Ferror handling, audit\u002Fprovenance, and explicit user confirmation. The model never becomes the authority for permissions or side effects.\n8. Store generated drafts, observations, predictions, transcripts, translations, and classifications separately from trusted domain truth. Show source context\u002Funcertainty when it matters, allow edit\u002Freject\u002Fretry, redact logs, and define raw-input\u002Foutput retention and deletion.\n9. Evaluate representative, empty, oversized, multilingual, adversarial, safety-sensitive, low-quality, and stale-input fixtures. Record quality, abstention\u002Frefusal, correction rate, latency, memory, battery, thermal state, dropped inputs, and device\u002FOS\u002Fmodel configuration.\n10. Verify the actual physical device and target language\u002Fmodel configuration for any claim that depends on Apple Intelligence availability, camera\u002Fmicrophone\u002Fsensor behavior, on-device performance, or system-surface invocation. A simulator, mock, preview, or successful compile is narrower evidence.\n\n## Fast path\n\nEstablish a deterministic baseline before tuning a model. Compare a representative fixture set covering valid output, malformed output, refusal, missing context, and oversized context, and promote only fields that pass schema, safety, and human-review gates. Expand the fixture set only when a failure reveals a new risk class.\n\n## Evaluation and safety contract\n\nFor each evaluation record:\n\n- route\u002Fframework\u002FAPI and source URL;\n- target SDK, OS\u002Fbuild, device family, region\u002Flanguage, Apple Intelligence\u002Fmodel\u002Fasset state, and model\u002Fprompt\u002Fschema version;\n- input source, redaction\u002Fminimization, maximum size, retention, and consent\u002Fpermission state;\n- expected behavior, fixture class, output type, quality\u002Fcorrection\u002Fabstention metric, latency, memory, battery, and thermal observations;\n- validation\u002Freview\u002Fcommit rule, fallback, and unresolved risks.\n\nRe-run representative prompts\u002Ffixtures after OS, SDK, model, language, prompt, schema, preprocessing, or tool changes. Test prompt-injection-like content when user or web content enters the context. Apple guardrails are a base layer; add app-specific domain, audience, privacy, and harm validation.\n\n## Non-negotiable safety and evidence rules\n\n- Never present generated text, OCR, classification, translation, transcription, sound analysis, or model output as medical\u002Flegal\u002Ffinancial truth, identity, authenticity, guaranteed accuracy, or authorization without domain validation and appropriate human review.\n- Never claim “on-device,” “private,” “real-time,” “supported,” “accurate,” or “safe” from an imported framework, model response, simulator, preview, mock, one prompt, one language, one device, or successful compile.\n- Keep source permission, model availability, input availability, output quality, domain validation, and side-effect authorization separate. A model that is ready can still return an unusable or unsafe proposal.\n- Do not send sensitive prompts, audio, images, health\u002Fpersonal data, or model output remotely or retain raw inputs beyond the stated feature. Do not add broad permissions, analytics, telemetry, accounts, or servers merely to make an AI demo work.\n- Cancel and clean up model sessions, capture, inference, translation, transcription, tool calls, and background tasks when the feature ends. Ignore stale results so an older proposal cannot overwrite newer user state.\n\nPer-framework boundaries, deliverable, change boundary, and sources: `references\u002Fdetails.md`.\n",{"name":5,"description":7,"allowed-tools":24,"last_verified":25},"Read, Grep, Glob, Edit, Write, Bash",1787097600000,"main","f846d9a97082ff30d613b55a085cef425d822473","plugins\u002Fios-ops\u002Fskills\u002Fios-on-device-intelligence","https:\u002F\u002Fgithub.com\u002Fshotcowboystyle\u002Fag-skills","https:\u002F\u002Fgithub.com\u002Fshotcowboystyle","https:\u002F\u002Fshotcowboystyle.github.io","Building smarter with AI, automation & modern tools. Exploring how far one person can go without a big team using today’s AI.","2026-11-10 04:58:27",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":46},[36,39,42],{"tool":37,"command":38},"npx","npx skills add shotcowboystyle\u002Fag-skills --skill ios-on-device-intelligence",{"tool":40,"command":41},"bunx","bunx skills add shotcowboystyle\u002Fag-skills --skill ios-on-device-intelligence",{"tool":43,"command":44},"pnpm","pnpm dlx skills add shotcowboystyle\u002Fag-skills --skill ios-on-device-intelligence","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fshotcowboystyle\u002Fag-skills\u002Fblob\u002Fmain\u002Fplugins\u002Fios-ops\u002Fskills\u002Fios-on-device-intelligence\u002FSKILL.md",true,[48],{"path":49,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":52,"sourceHtmlUrl":53},"references\u002Fdetails.md","REFERENCE","参考文档",8108,"https:\u002F\u002Fgithub.com\u002Fshotcowboystyle\u002Fag-skills\u002Fblob\u002Fmain\u002Fplugins\u002Fios-ops\u002Fskills\u002Fios-on-device-intelligence\u002Freferences\u002Fdetails.md",[55,61,68,75,82,89],{"skillNo":56,"name":57,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":58,"categories":59,"keywords":60,"publishedAt":19},"01M7F1YRY5S5QRJ1V0BJS130QX","no-description","https:\u002F\u002Fgithub.com\u002Fshotcowboystyle\u002Fag-skills\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fno-description\u002FSKILL.md",[],[],{"skillNo":62,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":65,"categories":66,"keywords":67,"publishedAt":19},"01M7F1Y9BV2FJNQJ3NB32JS5CM","good-skill","Fixture skill that satisfies every rule.","https:\u002F\u002Fgithub.com\u002Fshotcowboystyle\u002Fag-skills\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fgood-skill\u002FSKILL.md",[],[],{"skillNo":69,"name":70,"nameZh":6,"description":71,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":72,"categories":73,"keywords":74,"publishedAt":19},"01M7F1XXBQ3NQRXZSS0KAGAR9V","broken-link","Links to a reference file that does not exist.","https:\u002F\u002Fgithub.com\u002Fshotcowboystyle\u002Fag-skills\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fbroken-link\u002FSKILL.md",[],[],{"skillNo":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":19},"01M7F1XBX4R46NR9W0HXQ7XCDX","Bad_Name","Name violates the kebab-case rule and does not match the directory.","https:\u002F\u002Fgithub.com\u002Fshotcowboystyle\u002Fag-skills\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fbad-name\u002FSKILL.md",[],[],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":19},"01M7F1WNAF8WNEQ7S5HKT5VK0D","skill-authoring","Write, review, and validate portable agent skills in SKILL.md format. Use when creating a new skill, editing an existing one, or asked whether a skill follows the Agent Skills spec.","https:\u002F\u002Fgithub.com\u002Fshotcowboystyle\u002Fag-skills\u002Fblob\u002Fmain\u002Fskills\u002Fskill-authoring\u002FSKILL.md",[],[],{"skillNo":90,"name":91,"nameZh":6,"description":92,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":93,"categories":94,"keywords":95,"publishedAt":19},"01M7F1W30TC0XFNSF7YBRGETK2","weigh-bags","Record scale readings for each bag — empty (tare), outbound, current, final — and reconcile them against the itemised inventory to find unaccounted weight. Use when the user gives a bag weight, says how heavy something is, weighs their luggage, or asks how close to the limit they are. Appends to bags.yaml and runs scripts\u002Freconcile.py.","https:\u002F\u002Fgithub.com\u002Fshotcowboystyle\u002Fag-skills\u002Fblob\u002Fmain\u002Fplugins\u002Ftravel-packing-assistant\u002Fskills\u002Fweigh-bags\u002FSKILL.md",[],[]]