Analyze and search photo libraries using local computer vision without cloud uploads. Built for developers who need to index, tag, and query images programmatically with privacy-first processing.
io.github.prasadabhishek/photographi-mcp
Local install
STDIO
No auth required
How models use it and what it is built for.
Analyze and search photo libraries using local computer vision without cloud uploads. Built for developers who need to index, tag, and query images programmatically with privacy-first processing.
Local install — runs as a subprocess.
Where to find authoritative docs and source for Prasadabhishek Photographi.
Paste any of these into Agent Studio after connecting Prasadabhishek Photographi.
Common questions about connecting and running Prasadabhishek Photographi.
Does photographi-mcp upload images to the cloud?
No. It's a local computer vision engine that processes images on your machine, keeping all photo data private and under your control.
What image formats and photo library structures does it support?
The registry describes it as a visual intelligence engine for photo libraries, but specific format and library compatibility details are in the full documentation. Check the project repository for supported formats and integration examples.
How do I install and run photographi-mcp?
Install via `uvx photographi-mcp@0.2.6`. It runs over stdio transport, making it compatible with MCP clients that support local tool execution.
Can I use this to build a searchable photo app?
Yes. The MCP interface lets you query and tag images programmatically, making it suitable for building custom photo search, organization, and discovery tools.
What's the performance impact of running local computer vision on large libraries?
Local processing trades cloud latency for local compute time. Performance depends on your hardware and library size; refer to the project docs for benchmarks and optimization tips.
MCP Playground runs 10,000+ hosted MCP servers — GitHub, Linear, Notion, Stripe, Sentry and more — across Claude, GPT, Gemini, DeepSeek and 30+ AI models. Compare model answers side-by-side, save agent presets, share runs. Zero install.
Open Agent Studio