"""
š Welcome to your Smithery project!
To run your server, use "uv run dev"
To test interactively, use "uv run playground"
You might find this resources useful:
š§āš» MCP's Python SDK (helps you define your server)
https://github.com/modelcontextprotocol/python-sdk
"""
from mcp.server.fastmcp import Context, FastMCP
from pydantic import BaseModel, Field
from smithery.decorators import smithery
# Optional: If you want to receive session-level config from user, define it here
class ConfigSchema(BaseModel):
# access_token: str = Field(..., description="Your access token for authentication")
pirate_mode: bool = Field(False, description="Speak like a pirate")
num_papers: int = Field(5, description="Number of arXiv papers to return")
# For servers with configuration:
@smithery.server(config_schema=ConfigSchema)
# For servers without configuration, simply use:
# @smithery.server()
def create_server():
"""Create and configure the MCP server."""
# Create your FastMCP server as usual
server = FastMCP("Say Hello and search arXiv")
# Add a tool
@server.tool()
def hello(name: str, ctx: Context) -> str:
"""Say hello to someone."""
# Access session-specific config through context
session_config = ctx.session_config
# In real apps, use token for API requests:
# requests.get(url, headers={"Authorization": f"Bearer {session_config.access_token}"})
# if not session_config.access_token:
# return "Error: Access token required"
# Create greeting based on pirate mode
if session_config.pirate_mode:
return f"Ahoy, {name}!"
else:
return f"Hello, {name}!"
# Tool: Find new arXiv papers in a category from the last day (via RSS)
@server.tool()
def fetch_current_arxiv_postings_rss(category: str, ctx: Context) -> list:
"""
Returns all of today's brand-new arXiv submissions in a category via RSS.
"""
import feedparser
url = f"https://rss.arxiv.org/rss/{category}"
feed = feedparser.parse(url)
results = []
for e in feed.entries:
announce_type = getattr(e, "arxiv_announce_type", None)
if "replace" in announce_type.lower():
continue # skip replaced articles but keeps new cross-lists
results.append({
"title": e.title,
"authors": e.get("authors", []),
"summary": e.summary,
"url": e.link,
"published": str(e.published) if hasattr(e, "published") else None,
})
return results
# Tool: Find new arXiv papers in a category from the last day (via RSS), filtered by keyword
@server.tool()
def keyword_search_arxiv_rss(category: str, keyword: str, ctx: Context) -> list:
"""Search the daily arXiv RSS feed for a category, filtering by keyword in title or summary."""
import feedparser
keyword_lower = keyword.lower()
url = f"https://rss.arxiv.org/rss/{category}"
feed = feedparser.parse(url)
results = []
for e in feed.entries:
announce_type = getattr(e, "arxiv_announce_type", None)
if "replace" in announce_type.lower():
continue # skip replaced articles but keeps cross-lists
text = (e.title + " " + e.summary).lower()
if keyword_lower in text:
results.append({
"title": e.title,
"authors": e.get("authors", []), # RSS usually provides this
"summary": e.summary,
"url": e.link,
"published": str(e.published) if hasattr(e, "published") else None,
})
return results
# Add a resource
@server.resource("history://hello-world")
def hello_world() -> str:
"""The origin story of the famous 'Hello, World' program."""
return (
'"Hello, World" first appeared in a 1972 Bell Labs memo by '
"Brian Kernighan and later became the iconic first program "
"for beginners in countless languages."
)
# Add a prompt
@server.prompt()
def greet(name: str) -> list:
"""Generate a greeting prompt."""
return [
{
"role": "user",
"content": f"Say hello to {name}",
},
]
return server
