Apple's Foundation Models framework is still under active development
(built against macOS 27 beta) and likely to stay rough for a while —
small 4K context window, occasional generation errors, no tool
support yet. Surface that clearly wherever it appears:
- Model name: "Apple On-Device (Beta)" (shows in header/model picker)
- Model description: notes the beta status and what to expect
- Settings -> General: "⚠️ Beta — ..." disclaimer under the Apple
Intelligence section, same style as the existing Paperless-NGX beta
note
- Credits panel: same disclaimer for the Apple Intelligence entry
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Apple Intelligence / Foundation Models is now genuinely available on
this machine (macOS 27 beta 4) — it was reverted back in June
(f63226b) when it wasn't. Ports the reverted AppleFoundationProvider
forward onto 2.4.3, adapted for everything that's changed since:
PolyForm license headers, current AIProvider protocol shape, current
Settings.Provider/ProviderRegistry/CreditsView/SettingsView structure.
Fixes a real bug found via live testing: LanguageModelSession.
GenerationError was deprecated in macOS 27.0 in favor of a new
LanguageModelError type. On a macOS 27+ runtime, generation failures
now throw LanguageModelError, not GenerationError, so the original
error-mapping catch never matched and Apple's raw error text leaked
to the user instead of oAI's friendly message. Now dispatches to
whichever type the runtime actually throws, gated with
@available(macOS 27.0, *), keeping the old GenerationError path as
a fallback for macOS 26.x (the app's actual deployment target).
Confirmed end-to-end in the live app: provider selectable, Settings
shows a live "Available" badge, chat header shows correct branding,
and — a real, expected Phase 1 limitation — oAI's default system
prompt (active Agent Skills + MCP tool guidance, ~16K tokens on this
machine) exceeds the on-device model's 4K context window on the
very first message. The friendly error message now correctly reports
this instead of Apple's raw string. Tool calling remains out of scope
until Phase 3.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
The oAITests PBXNativeTarget (PBXContainerItemProxy, PBXTargetDependency,
XCBuildConfiguration with TEST_HOST/BUNDLE_LOADER, the "recommended
settings" changes accepted when the target was created -- DEAD_CODE_STRIPPING,
CLANG_ANALYZER_LOCALIZABILITY_NONLOCALIZED, STRING_CATALOG_GENERATE_SYMBOLS,
DEVELOPMENT_TEAM moved to project-level inheritance) was somehow never
actually staged in the very first "Add real oAITests target" commit
(8c7fb59) despite every xcodebuild test run since then depending on it
being present on disk. Every subsequent commit this session only staged
specific file paths (never oAI.xcodeproj again), so the gap went
unnoticed until a full `git status` review here.
Without this, anyone else pulling the branch (or a truly clean checkout
on this machine) would have all the .swift test files but no target to
compile them into -- xcodebuild test would fail to find oAITests at all.
Confirmed the diff is exactly the expected target-wiring content, nothing
unrelated or corrupted, before committing.
Every change here is either dropping `private` (still invisible
outside the module, @testable import just needs internal-or-wider)
or converting a self-independent instance method to `static func`
(ChatViewModel.inferProvider/calculateCost/detectGoodbyePhrase --
none of the three ever touched `self`, and constructing a real
ChatViewModel triggers a real network call in init, so static-ifying
them sidesteps that entirely rather than fighting it). Call sites
updated to `Self.foo(...)` where the static conversion required it.
Touches: GitSyncService's URL/secret-scanning helpers,
EmbeddingService's embedding (de)serialization, ContextSelectionService's
importance scoring, and the request-building/response-parsing helpers
on all four providers (OpenRouter, Ollama, OpenAI, Anthropic) -- the
core AI request/response layer, previously 100% untested.
Verified: full existing test suite (27 tests) still green, no
regressions. Tests for these functions land in the next commit.
Phase 2 of the test-suite rollout plan (peaceful-baking-kurzweil).
Consolidates the mac.oai.pm subdomain references (introduced in the
PolyForm Noncommercial relicense) to the root oai.pm domain, across
the LICENSE file, README, and all Swift source file headers.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Switches the project from AGPL to a source-available license that
restricts commercial use — selling oAI or any part of it, standalone
or bundled into another product/service, now requires a separate
commercial license from the copyright holder. Noncommercial use,
study, modification, and sharing remain fully permitted.
Updates: LICENSE (canonical PolyForm Noncommercial 1.0.0 text +
commercial licensing contact note), SPDX headers and file-header
boilerplate across all Swift source files, the in-app About dialog's
license link (+ its localization catalog entry), README.md and
DEVELOPMENT.md license sections.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Fetches /api/v1/images/models in parallel with /models and merges results
into the model picker. Image-only models (e.g. Sourceful, Seedream, Flux
via this endpoint) were previously invisible since they don't appear in the
standard /models endpoint.
Models from the images API get usesImagesAPI=true and route through a new
generateImageAPIResponse() path in ChatViewModel that POSTs to /api/v1/images
with {model, prompt} instead of the chat completions endpoint. The response's
b64_json data is decoded and displayed via the existing GeneratedImagesView.
Cost is taken directly from the usage.cost field in the images API response
(USD per image) via a new rawCostUSD field on ChatResponse.Usage, bypassing
the token-based calculateCost() path used for chat models.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Personal Data Tools: native Calendar, Reminders, Contacts (hidden pending
Apple TCC fix in beta), and Location & Maps access via EventKit, Contacts
framework, and MapKit. Write actions (create event/reminder, complete
reminder) gate through an approval sheet. Four hardened-runtime entitlements
added to oAI.entitlements; Info.plist usage strings added for all services.
Personal Data section shows a β badge while Contacts is hidden.
2nd Brain always-trust: inline toggle on the Agent Skills row for the skill
named "2nd Brain" skips the bash approval dialog when the command contains
.brain_helper.py, gated by three runtime checks in MCPService.
Research agents: spawn_research_agents tool runs up to 5 concurrent read-only
sub-agents (read_file, list_directory, search_files, web_search — no write,
no bash, no nesting). Bounded by maxConcurrentAgents setting (default 3) and
a hard ceiling of 8 tasks. Added items field to Tool.Function.Parameters.Property
for JSON Schema array support; wired into AnthropicProvider.convertParametersToDict.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Caches the system prompt/tools and growing conversation history via
cache_control breakpoints, cutting cost and latency on repeated turns.
Covers both the regular chat path and the tool-calling loop
(chatWithToolMessages), which has its own request-building code and was
initially missed. Cost calculation now accounts for cache write/read
pricing instead of treating all input tokens as full price. Verified
live: cache reads grow turn-over-turn in oAI.log.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- New AppleFoundationProvider using FoundationModels framework (macOS 27+)
- Streaming via streamResponse(to:) → ResponseStream<String> snapshot deltas
- Session built with system prompt + conversation history injected as instructions text
- Full error mapping: context exceeded, guardrail violation, rate limit, availability states
- Settings.Provider.appleOnDevice case wired through ProviderRegistry, Color+Extensions, CreditsView
- inferProvider() detects "apple-" prefix model IDs
- Settings → General: Apple Intelligence section with live availability badge and deep link to System Settings
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Jarvis integration: manage oAI-Web agents and usage from inside the app (/jarvis command, Settings tab 11)
- Model category filter: keyword-based categorisation with popover picker in model selector
- Categories shown in ModelInfoView with coloured chips; dot indicators on model rows
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
runeandClaude Sonnet 5