ProWorkBench is unclaimed —

ProWorkBench

Local-first AI workbench with approvals and execution control.·proworkbench.com

ProWorkBench is a controlled AI workspace that developers install and run on their own machine. Users bring their own OpenAI-compatible backend, manage Model Context Protocol (MCP) servers, and must approve every tool call before it executes. It provides an execution layer with explicit approvals, tool governance, and local logging rather than a hosted, autonomous AI assistant.

What it's for

Setting up and managing local dev environments for AI workApproving or denying AI tool calls before executionManaging and configuring MCP servers from one placeRunning doctor checks to verify environment and backend healthReviewing local logs of conversations, tool calls, and approvals

Features 12

Approvals system requiring explicit confirmation before any tool call runs
Block-by-default tools with per-risk policy controls
Bring-your-own OpenAI-compatible backend support
Canvas that stores outputs only and never executes actions
Doctor checks to verify environment, backend connection, and tool health
Local state and logging via a local SQLite database
MCP server manager to add, configure, enable, and disable MCP servers
Native installers for Windows, macOS, and Linux
No hosted dashboard or server-side account required
Telegram and Slack integration restricted to chat/inbox/notifications only, with tool/MCP execution hard-blocked from those channels
Tool governance dashboard showing every loaded tool, its source, and permissions
Unified approval queue for tool runs and MCP lifecycle actions

At a glance

TypeDesktop app

DeploymentInstalled (local)

PlatformsWindows, macOS, Linux

Fordevelopers, operators

CompanyProWorkBench

Integrations

llama.cppOllamaAnthropicTelegramLM StudiovLLMText Generation WebUIOpenAISlack

Resources

Docs

Pricing

A one-time license is offered with instant download after purchase and a 30-day money-back guarantee. Specific pricing is not publicly listed.