edgeever, developed by Tianma If, is an open-source knowledge base that revives Evernote-style note-taking for self-hosted, AI-connected workflows. The app converts static notes into a queryable resource for AI agents, offering full-text search, a classic three-pane editor, and native Model Context Protocol endpoints for assistant access. It supports ENEX migration, a REST API with OpenAPI schema, PWA access, and Cloudflare or Docker deployment. Long-time Evernote users and AI power users gain stronger data ownership and agent integration.

The app repurposes archived notes into a queryable context store for agents

Designed as an AI-native workspace, the tool exposes stored entries so external assistants can request relevant passages rather than reading raw files. That approach converts existing collections into addressable context the model can reference during generation. Migration scripts accept Evernote ENEX exports, enabling large archives to become searchable material, and the three-pane interface preserves a familiar navigation pattern for people moving from older note platforms.

MCP endpoints make assistant access practical, but outputs require human checking

The built-in Model Context Protocol server lets compatible clients fetch structured note context; named compatible clients include Claude Desktop and Cursor. This arrangement supplies the model with targeted text to synthesize answers, yet generated responses reflect the model's training patterns and are not independently validated. Verify assistant output against original notes or source material before using it for decision-making, technical research, or other high-stakes tasks.

Serverless and container options lower hosting barriers while requiring setup knowledge

The developer optimizes deployment for serverless Cloudflare components as well as Docker, so users avoid maintaining a classic VPS but must configure platform services or containers. The app exposes a REST API and an OpenAPI schema for automation, and PWA support provides an app-like browser experience on desktop and mobile. Expect initial configuration steps for storage bindings and database choices when bringing up an instance.

Open data architecture favors developers and scripting-heavy workflows

The app's open-source codebase and open-data design let third-party tools ingest, index, and enrich notes programmatically. The MCP server is compatible with MCP-compliant clients, creating a pipeline for assistants and coding tools. Developers can automate ENEX imports, call the REST endpoints from scripts, and add custom connectors. For anyone integrating multiple agent clients or bespoke tooling, this architecture supports programmatic context retrieval and pipeline customization.

Best suited to technically minded users who want control over notes and agent access

The app is a practical choice for people who accept hands-on deployment and intend to pair their knowledge base with assistant tools. Expect to perform platform setup and to treat assistant-generated answers as drafts that need verification. Users seeking a plug-and-play consumer client without any configuration work may find the initial onboarding demanding, while those who value control and extensibility will find it fit for purpose.