Local Memory Sidekick

No-login custom tracker with an MCP server so your local AI can log and review for you

Problem

Many people want to track custom metrics (habits, project progress, ideas), but existing tools are built around predefined categories. A Show HN project offers custom tracking with an MCP server, showing users want to manage personal data through their favorite AI instead of being locked into an app.

Solution

A no-login, local-first tracker where users define their own variables and recording methods. It also exposes an MCP server so local AI (e.g., Claude Desktop) can read/write data, generate briefings, and guides. Data stays on the user's machine or their own cloud drive.

Users

Tool-tinkering individuals, heavy AI users, quantified-self enthusiasts.

MVP

No-login local web app; custom variable and record forms; basic stats and trend charts; MCP server interface; one-click data export.

Revenue

Free local version plus paid cloud sync or team tier; subscription for advanced AI features via MCP integration.

Why now

The MCP ecosystem is nascent, and users crave plugging personal data into their own AI workflows; an indie dev can quickly own this niche.