{"slug": "stop-wasting-tokens-re-uploading-screenshots-and-specs-my-mcp-setup", "title": "Stop wasting tokens re-uploading screenshots and specs: My MCP setup", "summary": "A developer built Dokpod.io, an AI knowledge vault that uses the Model Context Protocol (MCP) to let AI coding assistants like Cursor and Claude Code query indexed project assets on demand, eliminating the need to re-upload screenshots and specs in every session. The tool aims to reduce token consumption and context clutter by indexing media and documents once and providing an MCP connection for retrieval.", "body_md": "If you use Cursor or Claude Code heavily, you probably know this workflow: You start a new session, drag and drop a bunch of UI screenshots, architecture diagrams, or heavy project specs into the chat, and tell the AI to look at this.\n\nIt works, but it causes two massive problems:\n\nToken Burn (and Credit Drain): Vision tokens and heavy text files are expensive. You waste your API credits processing those same screenshots and docs every single time you spin up a new chat.\n\nContext Clutter: The AI's context window gets clogged. Its logic degrades because it’s carrying all that heavy media and text around in its short term memory.\n\nI got tired of burning through my API credits on this daily, so I started looking into the Model Context Protocol (MCP).\n\nWhy MCP is the answer\n\nInstead of dumping static files and images directly into the prompt, MCP allows your AI editor to query a local or remote server only when it needs specific information.\n\nThink of it like giving Cursor a direct database connection to your project's assets. It indexes the data once, and the AI retrieves just the pieces it needs to answer your specific coding question. The token savings are ridiculous.\n\nHow I automated this (Building Dokpod)\n\nYou can build a local MCP server yourself, but managing the indexing for mixed media (images, video walkthroughs, and text), handling local environments, and keeping connections stable became its own headache.\n\nSo, I built [Dokpod.io] to automate the entire thing.\n\nIt acts as an AI knowledge vault. You upload your UI screenshots, video walkthroughs, API docs, and codebase context into Dokpod once. It handles the indexing and gives you a simple MCP connection to plug straight into Cursor or Claude.\n\nThe result:\n\nZero repetitive uploading for images, videos, or text.\n\nMassive reduction in input tokens (saving your credits and limits).\n\nThe AI actually remembers your UI references and architecture across different coding sessions.\n\nI need your technical feedback\n\nIf you are wrestling with context limits or burning through your credits, I’d love for you to try this workflow out. You can check out the vault at [Dokpod.io](https://dokpod.io).\n\nI'm currently working on improving the indexing speed for heavier media and adding more native integrations. If you give it a spin in your setup, please tear it apart and let me know what I should build next or what I can improve.\n\nLet me know what you think in the comments!", "url": "https://wpnews.pro/news/stop-wasting-tokens-re-uploading-screenshots-and-specs-my-mcp-setup", "canonical_source": "https://dev.to/n0tzayn/stop-wasting-tokens-re-uploading-screenshots-and-specs-my-mcp-setup-cgd", "published_at": "2026-09-02 18:47:39+00:00", "updated_at": "2026-09-02 18:53:38.244088+00:00", "lang": "en", "topics": ["developer-tools", "ai-tools", "ai-infrastructure"], "entities": ["Dokpod.io", "Cursor", "Claude Code", "Model Context Protocol"], "alternates": {"html": "https://wpnews.pro/news/stop-wasting-tokens-re-uploading-screenshots-and-specs-my-mcp-setup", "markdown": "https://wpnews.pro/news/stop-wasting-tokens-re-uploading-screenshots-and-specs-my-mcp-setup.md", "text": "https://wpnews.pro/news/stop-wasting-tokens-re-uploading-screenshots-and-specs-my-mcp-setup.txt", "jsonld": "https://wpnews.pro/news/stop-wasting-tokens-re-uploading-screenshots-and-specs-my-mcp-setup.jsonld"}}