Simple Commenter AI

Product Introduction

1. Definition: Simple Commenter AI is a Model Context Protocol (MCP) server and workflow automation platform designed for web developers and agencies. It acts as a specialized middleware that connects client website feedback directly to AI coding assistants like ChatGPT, Claude, Cursor, Claude Code, and Codex.
2. Core Value Proposition: It exists to streamline the website feedback-to-fix lifecycle by providing AI agents with the rich, contextual data they need to propose accurate solutions. It eliminates manual context-switching by automatically delivering client comments, screenshots, element details, and JavaScript error logs to the developer's AI tool of choice, enabling faster, more informed code fixes.

Main Features

1. Unified Feedback Context Delivery: The system captures and packages every piece of feedback with comprehensive technical metadata. When a client leaves a comment, the MCP server provides the AI with the exact page URL, browser and OS data, a screenshot, the specific HTML/CSS selector (e.g., .hero h1), and captured console errors. This creates a complete "dev brief" for the AI to analyze.
2. Flexible MCP Integration: Supports both hosted (OAuth) and local (stdio) MCP server connections. For tools like ChatGPT and Claude, it offers a secure, hosted OAuth endpoint (https://www.simplecommenter.com/api/mcp) with granular project-level permissions. For local IDEs like Cursor and Claude Code, it provides an npm package (@simple-commenter/mcp-server) for stdio-based integration, initialized via a command-line wizard (npx @simple-commenter/mcp-server init).
3. AI-Actionable Workflow Tools: Beyond reading feedback, the MCP server grants authorized AI agents the ability to take action within the Simple Commenter platform. This includes posting replies to clients, updating comment statuses (e.g., from "To Do" to "Review"), changing priorities, and generating project summary reports, all contingent on the permissions granted during OAuth setup.

Problems Solved

1. Pain Point: The inefficient, manual triage of vague client feedback. Developers waste time asking for clarifications, replicating browser/device issues, and hunting for the relevant code section, a process often described as "context switching hell."
2. Target Audience: Freelance web developers, digital agencies, SaaS product teams, and in-house web development departments who manage client websites and need to process visual and functional feedback efficiently.
3. Use Cases: A client reports a layout bug on mobile. Instead of a back-and-forth, the developer asks their connected AI in Cursor to "investigate comment #913." The AI instantly pulls the screenshot, sees the .hero h1 overflow at 375px, analyzes the relevant CSS file, and proposes a responsive fix using clamp(). The developer reviews and tests the code before asking the AI to post a resolution reply.

Unique Advantages

1. Differentiation: Unlike generic screen capture or bug-tracking tools, Simple Commenter AI is built specifically as a conduit for AI agents. It doesn't just store feedback; it structures it in a way that large language models (LLMs) can consume and act upon directly, bridging the gap between client communication and AI-powered development environments.
2. Key Innovation: Its implementation as a standardized MCP (Model Context Protocol) server. This protocol, pioneered by Anthropic, allows it to connect seamlessly to a growing ecosystem of AI tools without requiring custom plugins for each one. This standards-based approach ensures future compatibility and avoids vendor lock-in to a single AI platform.

Frequently Asked Questions (FAQ)

1. How does Simple Commenter AI connect to ChatGPT or Claude? It uses the Model Context Protocol (MCP) via a hosted server URL. You enable developer mode in ChatGPT or add a custom connector in Claude, input the Simple Commenter MCP endpoint (https://www.simplecommenter.com/api/mcp), and complete an OAuth sign-in flow to authorize project access.
2. Can my AI automatically push code changes to my website? No. Simple Commenter AI is a feedback and context delivery system. It provides the diagnostic information to your AI coding assistant. The AI still requires separate access to your code repository (e.g., via GitHub) to propose fixes. All code changes must be reviewed, tested, and deployed through your standard development workflow.
3. Is the MCP connection free, and do clients need an AI account? The MCP server integration is included with all paid Simple Commenter plans and the 14-day trial. There is no additional subscription fee from Simple Commenter. Your clients do not need any AI accounts; they interact only with the Simple Commenter feedback widget as usual.
4. What's the difference between the hosted OAuth and local npm setup? The hosted OAuth connection (for ChatGPT, Claude, Codex) allows you to select specific projects and permissions (read/write) during sign-in. The local npm setup (for Cursor, Claude Code) uses a terminal-based login that grants account-wide access, making it ideal for developers working across all their projects in a local IDE.