About Kinsy
Kinsy is a remote MCP server for assistants and agents that creates editable PowerPoint decks from a brief and source material. It exposes deck creation, progress, status, export, slide editing, and theme tools through the caller's Kinsy API key.
Key features
- 01Create decks from briefs and source text
- 02Stream deck-generation progress
- 03Check status, links, and unsupported claims
- 04Export PPTX, PDF, or slide-image ZIP
- 05Rewrite individual slides
- 06List available themes
Use cases
- 01Turn a project brief and documents into a deck
- 02Track generation until a deck is ready
- 03Revise a slide using an instruction
- 04Download a deck as PPTX or PDF
MCP support
Kinsy exposes tools for creating, checking, editing, and exporting PowerPoint decks to MCP clients. Calls use the client's own Kinsy API key and the server stores no credentials.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://mcp.kinsy.ai/mcpclaude mcp add --transport http ai-kinsy https://mcp.kinsy.ai/mcp{"mcpServers":{"ai-kinsy":{"type":"http","url":"https://mcp.kinsy.ai/mcp"}}}Open in Cursor{"servers":{"ai-kinsy":{"type":"http","url":"https://mcp.kinsy.ai/mcp"}}}Open in VS Code{"mcpServers":{"ai-kinsy":{"type":"http","url":"https://mcp.kinsy.ai/mcp"}}}{"mcpServers":{"ai-kinsy":{"serverUrl":"https://mcp.kinsy.ai/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
https://mcp.kinsy.ai/mcpAuth requiredTools
Details
From the project READMEKinsy MCP server
The remote Model Context Protocol server behind https://mcp.kinsy.ai/mcp. Hosted assistants and agents connect to it with a Kinsy API key and turn a brief and the user's source material into a finished, editable PowerPoint (.pptx) deck.
It is a thin client over the public Kinsy API: every call is made with the caller's own key, so credits, limits and deck ownership are enforced by the API exactly as for the CLI. The server stores nothing and holds no credentials of its own.
Tools
| Tool | What it does | Changes anything? |
|---|---|---|
create_deck | Starts a deck from a brief and optional sources (text). Answers at once with a deck_id. | Spends credits |
wait_for_deck | Waits for a deck to finish (up to 540 s), streaming progress, then returns its links. | No |
get_deck | Status and links; optionally slide image links and the claims the sources could not back. | No |
export_deck | A fresh one-hour download link for the .pptx, PDF, or a zip of slide images. | No |
edit_slide | Rewrites one slide from an instruction. | Spends credits |
list_themes | The theme ids create_deck accepts. | No |
Results are a short sentence plus structured content. The server never echoes a brief or source text back, and never includes the key.
Connect
Create a key in Settings → API keys on kinsy.ai, then point your client at https://mcp.kinsy.ai/mcp with the header Authorization: Bearer ksy_live_…. Per-client instructions: kinsy.ai/docs/mcp.
Develop
python3 -m venv .venv && .venv/bin/pip install -r requirements.lock pytest
.venv/bin/python -m pytest
KINSY_MCP_SERVE=1 KINSY_API_BASE=https://kinsy.ai .venv/bin/uvicorn kinsy_mcp.server:app --port 8080
requirements.in lists the direct dependencies; scripts/lock.sh resolves the hash-pinned requirements.lock the image installs.
MIT. © Kinsy.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.