Chinvat: local MCP labor hub for governed multi-model task delegation
Chinvat is a free, open-source local MCP labor hub for Windows that allows any MCP-capable coordinator (e.g., Claude Code, Claude Desktop, Codex, Cursor, Hermes) to delegate work to local models, remote specialist models, Windows itself, and communication/publishing channels. It provides a persistent job queue, artifacts, and a policy layer that governs what crosses the bridge, solving the problem of managing multiple AI models and tools from a single governed interface.
Entities
Related
Local MCP labor hub for delegating work to models and tools
Chinvat is a free, open-source local MCP labor hub for Windows that provides a single server for MCP-capable coordinators (e.g., Claude Code, Claude Desktop, Codex, Cursor, Hermes) to delegate work to local models, remote specialist models, Windows itself, and communication/publishing channels. It includes a persistent job queue, artifacts, and a policy layer that decides what crosses the bridge, solving the problem of managing multiple connections and tools from a single interface.
Wigolo: MCP server giving LLM coding tools live web access with crawling and caching
Wigolo is an open-source MCP server that runs locally and provides LLM coding tools (Claude Code, Cursor, Codex, Gemini CLI) with ten web tools: multi-engine search with reranking, page fetching for JavaScript-heavy sites and PDFs, whole-docs-site crawling, structured extraction, and persistent local caching. It solves the problem of LLMs relying on stale training data by giving them real-time web access.
FableCut released: zero-dependency browser video editor with AI-driven timeline via MCP/REST
FableCut is a browser-based non-linear video editor that exposes its timeline as a JSON document, enabling AI agents to edit videos programmatically via MCP or REST APIs while the UI updates live. The tool has zero npm dependencies and is designed for integration with AI assistants like Claude Code.
Snitch: deterministic claim verifier for AI agent transcripts
Snitch is a tool that watches AI agent transcript files (Cursor, Claude Code, Codex, Pi, OpenCode) and verifies claims made in prose against actual evidence like tool calls, shell output, filesystem changes, git history, and session context. It uses deterministic regex patterns to extract claims and cross-references them, flagging inconsistencies. It helps developers trust their coding agents by catching when an agent's description doesn't match reality.
Sanbox: Isolated sandboxes for AI agents with MicroVM isolation
Sanbox is a platform that provides isolated, resumable sandboxes for running AI agents. It uses MicroVM isolation, a persistent filesystem, and a live trail of run events. The platform supports reusable templates, a CLI that works with Codex, Claude Code, Cursor, CI, or terminal, and can be self-hosted for security/compliance.