Self-hosted MCP access to visual feedback, recordings, diagnostics, project context, and verification workflows.
Copy the AI prompt to install this server into Claude Code, Cursor, or another agent — or use 1-click editor setup below.
One-click editor setup isn’t available for this listing yet — we don’t have a confirmed install command, and we’d rather show nothing than point your editor at the wrong package or host. Follow the project’s own setup instructions, linked above.
Inspect callable tools, capabilities, and parameters exposed to AI agents by Feedbacks.
Feedbacks MCP server gives coding assistants access to selected evidence from a Feedbacks team server. The surrounding application collects website feedback through screenshots, annotations, full-page captures, selected-text suggestions, video, and session-only recordings. It also stores discussion threads, event timelines, console data, network information, performance details, checksums, and verification checks.
The MCP connection is intended to add project and review context to agent work. Approved project instructions, reviewer guidance, and responsibilities can be provided through MCP, the API, and the CLI. Agent access is scoped rather than being an unrestricted view of the whole installation. Recording exports are also supported.
An administrator first deploys one Feedbacks installation for an organization. The team then connects the Chrome extension to that server and configures project context, members, and any optional GitHub App integration. Each developer configures a personal MCP connection using a scoped key. The extension captures browser evidence; it does not run the server.
Feedbacks MCP server works alongside the web application and its storage layer. Production deployments require HTTPS, PostgreSQL, and private S3-compatible storage. The repository also includes a CLI with a stdio MCP adapter, so an MCP-compatible coding assistant can connect through a local process after the application is available.
For local development, the documented sequence uses Node.js 22 or 24, npm, and Docker Compose for PostgreSQL:
npm ci..env.example to .env.docker compose -f compose.dev.yaml up -d --wait.BOOTSTRAP_EMAIL, BOOTSTRAP_NAME, and a unique BOOTSTRAP_PASSWORD before running bootstrap.npm run bootstrap, npm run build, and npm run dev:server.The bootstrap password should be removed after the first owner is created. A registry deployment is also documented through the public Docker Hub image and registry Compose configuration. Follow the project’s deployment documentation for production configuration, recovery, and storage fields.
The available MCP-related capabilities described by the project include:
The broader Feedbacks workflow supports assignments, categories, tags, saved views, point progress, and optional GitHub Issues. These application features are not necessarily separate MCP tools.
Feedbacks MCP server is part of an early 0.x release. One application deployment serves one organization. Public registration, billing, SSO, and shared-database multi-tenant customer support are not implemented. The team must provide and operate the infrastructure, including production storage and PostgreSQL.
The README names Claude Code, Codex, and Gemini CLI integrations. The repository includes bundled Claude Code and Codex plugins, while Gemini CLI uses a thin adapter to the same instructions. Access depends on the configured server, project scope, and developer credentials.
Factual signals from GitHub, npm, and our automated checks — not a rating.
No reviews yet — be the first to share how this listing worked for you.
Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.
[](https://allmcps.com/mcp/feedbacks)<a href="https://allmcps.com/mcp/feedbacks"><img src="https://allmcps.com/api/badge/feedbacks?style=directory" alt="Feedbacks on AllMCPs" /></a>