Uwe Chalas
Uwe Chalas is a German software developer and AI researcher whose public work focuses on AI-assisted development and local-first developer tools. His current public projects include NevoCortex, a cognitive AI assistant for Windows, and AiDex, an open-source MCP server that gives coding assistants persistent memory, structural code search and telemetry.
At a Glance
- Individual software developers
- AI-assisted development teams
- Users of MCP-compatible coding assistants
- Teams working across large or multiple codebases
- +1 more
AI Tools by Uwe Chalas
(1)
AiDex
MCP Server for Code Search
Discussions
No discussions yet
Be the first to start a discussion about Uwe Chalas
Latest News
AiDex v2.4.1 security release fixed command injection in aidex_screenshot and improved nested C/C++ build handling.
AiDex v2.4.0 added control push functionality and documented the dashboard/control API.
AiDex v2.3.0 added dashboard controls, Kotlin and Swift support, and multiple Viewer fixes.
AiDex v2.2.2 added Astro frontmatter support to the indexer.
Products & Services
An open-source, MIT-licensed MCP server for AI coding assistants. It provides persistent project and session memory, exact/semantic/hybrid code and documentation search, structural signatures and summaries, cross-project indexing, task management, screenshots, a universal LogHub and a live Viewer. It is distributed through GitHub and npm and requires Node.js 20+.
A cognitive AI assistant for Windows intended to understand what is happening on a user's PC and act as a personal expert system. The public concept combines real-time Windows UI Automation and OCR, curated system/application knowledge, persistent code indexing and an MCP server ecosystem; it is designed to be local, secure, GDPR-compliant and usable without a cloud connection.
Market Position
AiDex positions itself as a local-first, model-agnostic persistent memory and code-intelligence layer for MCP coding agents, emphasizing roughly 50x lower context usage than grep. Its practical alternatives include grep/ripgrep-based workflows and other code-indexing, semantic-search and AI coding-context tools; its differentiation is the combination of structural indexing, cross-session memory, cross-project search and live telemetry in one MCP server.
Founding Story
AiDex was created to address the context and continuity limits of conventional AI coding workflows: instead of repeatedly sending large files or relying on grep, the tool maintains a persistent local index and exposes compact, structure-aware results to an AI assistant. The project extends that idea with cross-session memory, semantic search and live telemetry.
Business Model
Revenue Model
AiDex is distributed as a public, MIT-licensed open-source npm package and GitHub project; its public package and repository describe installation and use but do not present paid plans.
Target Markets
- Individual software developers
- AI-assisted development teams
- Users of MCP-compatible coding assistants
- Teams working across large or multiple codebases
- Windows power users seeking local-first AI assistance
- Giving AI coding agents compact, structure-aware access to large codebases
- Maintaining memory and notes across coding sessions
- Semantic discovery of code, documentation and workspace knowledge
- Searching identifiers and changes across multiple repositories
- Monitoring applications through HTTP logs and live telemetry
- Providing AI-assisted Windows desktop understanding and automation through NevoCortex