Local-first AI assistant for notes, calendars, and daily priorities
Generates morning briefs by reading local markdown notebooks and system calendars to summarise scheduled events and daily priorities. Tallboy Consulting Inc supplies a desktop assistant for Windows that keeps data on-device. Key capabilities include markdown integration, multi-model backend support (Claude, OpenAI, or local LLMs), and system Calendar and Reminders access. It targets busy professionals, students, and parents who need context-aware planning without sending personal notes to remote servers.
Reads your files and builds persistent context
Yachito ingests plain-markdown notebooks and local system data, producing a persistent view of weekly priorities and long-term plans. The app operates on a local-first principle, so the assistant derives context from files stored on the machine rather than a remote corpus. That design lets the assistant reference specific notebook entries and calendar items when it generates summaries or suggestions.
System impact and runtime behaviour
The application requests Calendar and Reminders permissions to stay synchronized with schedules, which means it interacts directly with the desktop scheduling services. Because it supports local LLMs via Ollama and lets users supply external API keys, processing can run locally or through external models depending on configuration, affecting CPU and memory usage according to the chosen backend.
Privacy and safety controls are explicit
Yachito's local-first architecture keeps personal notes on-device and offers the option to run offline models, which provides data sovereignty for users unwilling to route content through third-party servers. The installer is signed and notarized, indicating code signing; the app still requires explicit system permissions, so users control which system data the assistant can read.
Interface and learning curve for productive use
The workflow expects users to keep notes in plain-markdown and to grant calendar access, which favors people comfortable organising files and adjusting model backends. The flexibility to use external API keys or a local model adds setup steps for non-technical users, while experienced users can tune the assistant by switching backends and pointing it at specific notebooks.
Practical for privacy-conscious power users, but gated by access
Yachito suits users who accept a short setup process to keep data on-device and who can manage markdown-based notebooks and model configuration. The current invite-only early access model limits immediate availability for wider teams. If you prioritise local data control and are prepared to configure a model backend, Yachito is a solid productivity companion. Recommended.
