Log inSubmit server
@skyelight MCP logo

@skyelight MCP

by @skyelight

Lets coding agents inspect Skyelight feedback, trace it to source elements, and update or resolve the resulting work.

InstallRepositoryWebsite
Full README Compare alternatives Embed widget Claim this listing

Install @skyelight MCP

Pick your client, then add it in one click or paste the config.

STDIONode.js
~/Library/Application Support/Claude/claude_desktop_config.json
claude_desktop_config.json
{
  "mcpServers": {
    "skyelight-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@skyelight/mcp"
      ],
      "env": {
        "SKYELIGHT_API_URL": "YOUR_VALUE_HERE",
        "SKYELIGHT_API_TOKEN": "YOUR_VALUE_HERE"
      }
    }
  }
}

Settings → Developer → Edit Config, merge this into mcpServers, then fully quit and reopen Claude Desktop.

Values are written into the snippet above so you can paste it as is. They never leave your browser.

Install not yet sandbox-tested— we haven't run this command through our automated sandbox yet. Not a red flag; we're working through the catalog.

Overview

@skyelight/mcp connects coding agents to Skyelight feedback from running applications, including discussion threads, pages, selected elements, source code, and images. Agents can inspect projects, search and review feedback, report progress, change status, assign work, and publish code or deployed applications for review. Use it when implementation work needs precise feedback context and a way to close the loop with reviewers. It supports remote OAuth connections and a local stdio mode authenticated with a personal token.

Use cases

  • Triage outstanding application feedback by page, status, assignee, or hand-off
  • Inspect the source context and visual anchor behind a reported issue
  • Reply to reviewers and move completed items to resolved
  • Publish a prototype or deployed application for Skyelight review

Key features

  • Reads threads, page context, anchors, code, and images
  • Searches feedback and detects similar or duplicate reports
  • Updates status, assignments, replies, rules, and decision context
  • Publishes code or deployed applications for review
  • Supports remote OAuth and local token authentication

Tools18

What @skyelight MCP lets your agent do. Expand a tool for its parameters.

Self-reported~317 tokens of context

Security & health

What it can reach and how reliably it runs — from our own checks, not the maintainer’s claims.

ActiveLast checked Today

Recent health check succeeded.

Supply-chain signal

No high-severity advisories surfaced by our automated scan.

Critical 0High 0Medium 0Low 0

Scanned 3h ago via OSV.dev · @skyelight/mcp (npm)

Adoption & maintenance

Factual signals from GitHub, npm, and our automated checks — not a rating.

GitHub stars1Stargazers on the source repository.
npm downloads1.6kPackage downloads in the last 30 days.
Last commit4d agoMost recent push to the default branch.
Tools exposed18Callable tools this server registers over MCP.

How @skyelight MCP works

What @skyelight/mcp does

@skyelight/mcp gives coding agents access to feedback collected in Skyelight. A feedback item can include its discussion, the page where it was created, the selected element or source location, related code, and images. Agents can use that context to understand a report before editing code, then reply on the original thread or update the item's workflow state.

The server also exposes project-level context. Agents can read workspace and project membership, review recent changes, retrieve decision history, save agreed rules, identify feedback associated with a file or component, and look for duplicate reports. This makes it suitable for coding-agent workflows that move from feedback triage to implementation and hand-off.

How it works

There are two connection modes. The remote mode uses HTTP and OAuth; the agent opens a browser for sign-in and approval, so a token is not stored locally. The local mode runs the package as an MCP stdio process and uses a Skyelight personal access token.

The tools are split between read and write operations. Reading includes workspace discovery, project listing, item search, item retrieval, member lookup, project reviews, source searches, duplicate searches, and decision-log access. Writing includes thread updates, new items, status changes, assignments, saved rules, merges, review-link publishing, and advancing the “last looked” bookmark used by whats_new. MCP annotations identify tool behavior, including the destructive nature of merge_items.

A local repository can contain .skyelight.json with a project ID and name. That lets local tools use the project by default. Without that binding, the agent selects a project through list_projects.

Setup and configuration

The shortest setup is:

Terminal
npx @skyelight/mcp init

The initializer detects supported coding agents, registers Skyelight's remote MCP endpoint, and leaves authentication to OAuth. Use --client cursor, --client all, or another supported client selection when needed. npx @skyelight/mcp remove removes the registered entry. SKYELIGHT_URL can point the initializer at another deployment.

For token-based use, create a personal access token in Skyelight under account settings and configure the server with SKYELIGHT_API_URL and SKYELIGHT_API_TOKEN. Credentials are checked from environment variables, .env.local, .env, and ~/.skyelight/credentials, in that order. The package can also be installed as a Claude Code plugin or a Gemini CLI extension, both of which use remote OAuth.

Tools and capabilities

The 18 tools cover:

  • Workspace, project, member, and project-review discovery.
  • Filtering, searching, opening, and tracking feedback items.
  • Thread replies, new findings, status changes, assignments, and duplicate merging.
  • Source-based feedback lookup, similarity checks, decision logs, and saved project rules.
  • Publishing a standalone HTML page, a React component with a default export, or a deployed application for review.

publish_review_link can host supported code on a Skyelight review URL or register an existing deployed URL and return the review link and script needed for the application.

Limitations and notes

The hosted-code publishing path accepts one HTML page or one React component with an export default; code that imports the rest of an application should be supplied as a deployed URL instead. A deployed application can be registered through its URL, including supported lovable.app and vercel.app addresses.

All plans expose the tools to owners, admins, and collaborators. Free plans allow 25 agent threads per month, while Pro and higher plans remove that limit. Reviewers can read and comment in the Skyelight web application but do not receive agent tools.

Read the full README View source on GitHub

Reviews

No reviews yet — be the first to share how it worked for you.

Frequently Asked Questions about @skyelight MCP

@skyelight/mcp is an MCP server that connects coding agents to Skyelight workspaces, projects, and application feedback. Its main tools list and search items, retrieve threads with page and source context, post updates, create findings, change statuses, assign work, and publish review links.

Run `npx @skyelight/mcp init` to register Skyelight's remote MCP server with a supported coding agent. The initializer waits for confirmation and uses OAuth during the first connection. For token-based local use, add `npx -y @skyelight/mcp` as a stdio server with the required Skyelight environment variables.

Yes, Cursor is supported by the initializer. Run `npx @skyelight/mcp init --client cursor` to register the remote server, then complete the browser-based OAuth sign-in when Cursor first connects.

Set `SKYELIGHT_API_URL` and `SKYELIGHT_API_TOKEN` in the MCP client's server configuration, then run the package with `npx -y @skyelight/mcp`. Create the token in Skyelight account settings under API Keys; it belongs to the organization where it was created.

Related MCP servers

All in Product Management All alternatives
Add a badge to your READMEShow it’s listed — live SVG, Markdown or HTML.

AllMCPs Directory Badge

Full Badge Customizer

Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.

Badge Style:
Live Dynamic SVG Preview@skyelight MCP AllMCPs Directory Badge
Markdown (GitHub README)
[![AllMCPs](https://allmcps.com/api/badge/skyelight-mcp?style=directory)](https://allmcps.com/mcp/skyelight-mcp)
HTML Embed
<a href="https://allmcps.com/mcp/skyelight-mcp"><img src="https://allmcps.com/api/badge/skyelight-mcp?style=directory" alt="@skyelight MCP on AllMCPs" /></a>