Changelog

Fix 1: Spending by Category charts changed from cramped side-by-side

grid to stacked vertical layout (bar 400px, pie 300px full-width).

Fix 2: Pie chart tooltip formatter now uses p.Model for the actual

chart point value instead of the captured closure variable, fixing

tooltips that could show incorrect dollar amounts.

New: Forecasts report tab (index 4) with three projection charts —

Spending Forecast, Income vs Expenses Forecast, and Net Worth

Projection. Uses weighted moving average (0.5/0.3/0.2 weights on

last 3 months) for spending and income/expense projections. Net

worth uses a 60/40 blend of linear trend and WMA for smoother

long-term projections. Forecast segments render with dashed stroke

and reduced fill opacity to visually distinguish from historical

data. Configurable forecast horizon: 3, 6, or 12 months.

Hardware profiling: GPU detection (CUDA/ROCm/Metal), CPU/SIMD capability

assessment (AVX2/AVX-512), and available memory detection. Composite

fitness scoring (0-100) with Green/Yellow/Red tiers determines local AI

viability and drives cloud provider recommendations.

New cloud providers: Mistral AI (EU-based, GDPR-native, HighPrivacy) and

Groq (fast inference, StandardApi). Both use OpenAiCompatibleProviderBase

which extracts shared OpenAI chat format logic.

Privacy tier system: PrivacyTier property on all providers — HighPrivacy

(Anthropic, Mistral, Local) vs StandardApi (OpenAI, Gemini, Groq).

Provider dropdown now shows "Privacy-First" labels.

Smart GPU offload: LocalLlamaProvider uses PlatformDetector.DetectGpu()

to set GpuLayerCount to -1 (all layers) when GPU is available, or 0

(CPU-only) when no accelerator is detected — prevents crashes and

unusable performance on CPU-only systems.

Settings UI: Hardware Report Card replaces single RAM recommendation line

showing RAM/GPU/CPU details, fitness score badge, and cloud-recommended

banner for Yellow/Red tier systems.

Desktop version bumped from 1.64.3 to 1.65.0.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Introduces PrivacyTier enum (HighPrivacy, StandardApi) to classify AI

providers by their data handling guarantees. Adds nullable PrivacyTier

property to AiProviderInfo record. Bumps SDK version from 1.64.0 to

1.65.0 to publish updated types before dependent desktop changes.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Extract and resolve wiki-links ([[type:id|Title]]) from user chat messages

via ILinkableItemProvider, injecting resolved entity metadata into the AI

system prompt so ACTION blocks can reference entities by ID. Track the last

successfully acted-on entity (ID + type) across the conversation so

pronouns like "that task" or "it" resolve correctly. Fields reset on new

chat. Bumps Desktop to 1.64.0.

Add unified settings overlay accessed via gear icon in toolbar.

TasksSettingsViewModel manages 4 tabs: Statuses (view/add/delete

status definitions with category and color), Fields (view/add/delete

custom field definitions with type selection), Board (select board,

view/add/delete columns, assign statuses), Rules (view/add/delete

automation rules with trigger type and enable/disable toggle).

Settings changes trigger task refresh on close. Updated

PLUGIN_CONTEXT.md and DetailedDescription to document all new

features from Phases 1-5.