Latchshot for Gemini CLI
Provides two MCP tools (capture_page and get_usage) that let Gemini CLI capture public web pages as PNG/JPEG/PDF and check render quota, all via a hosted Streamable HTTP endpoint without local browser.
README
Latchshot for Gemini CLI
Give Gemini CLI two bounded tools for public-page artifacts without installing a browser locally:
capture_pagereturns a PNG, JPEG, or PDF plus render diagnostics.get_usagereads the current plan and successful-render allowance without consuming quota.
The extension contains no executable code, hooks, commands, skills, agents, or always-on prompt content. It connects Gemini CLI directly to Latchshot's hosted MCP Streamable HTTP endpoint. Gemini CLI stores the API key as a sensitive extension setting and injects it only into the fixed authorization header.
Requirements
- Gemini CLI 0.51.0 or newer.
- A recurring Free-plan Latchshot key. The Free plan includes 100 successful renders per UTC calendar month and requires no credit card.
Install
Install the extension from its public GitHub repository:
gemini extensions install https://github.com/BaiqingL/latchshot-gemini-cli
Review the manifest when Gemini CLI asks for consent, then enter the Latchshot API key at the configuration prompt. The manifest marks the key as sensitive so Gemini CLI stores it through its secure extension-settings path. Do not pass --skip-settings.
Restart Gemini CLI, then run /extensions list or this terminal command:
gemini extensions list
The Latchshot server should expose capture_page and get_usage. To replace the key later, run:
gemini extensions config latchshot "Latchshot API key"
Example
Ask Gemini:
Use Latchshot to capture https://example.com as an 800 by 450 PNG. Report the byte count, render time, navigation state, fonts state, scripts state, and remaining quota.
A controlled integration check should use a disposable key and one public, non-sensitive URL. Artifact size and render time vary by page.
Boundaries
- Public HTTP and HTTPS targets on ports 80 and 443 only.
- Private, loopback, link-local, metadata, credential-bearing, mixed public/private DNS, and non-web-port targets are rejected.
- No cookies, sessions, arbitrary scripts, CAPTCHA solving, proxy rotation, or anti-bot bypass.
- The extension exposes no payment or plan-upgrade tool. Failed direct captures do not consume quota, and automatic overages are disabled.
- Target pages can observe a browser visit from Latchshot infrastructure. Capture only pages you are authorized to use.
Validate
npm test
npx --yes @google/gemini-cli@0.51.0 extensions validate .
The repository uses the gemini-cli-extension GitHub topic so it can be automatically indexed by Gemini CLI's daily gallery crawler after publication and validation. Direct installation from GitHub does not depend on gallery listing. No Google approval, affiliation, or gallery acceptance is claimed.
The repository is MIT licensed. The hosted Latchshot service remains a separate proprietary, quota-metered product.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.