pi-search
MCP server for pi that provides free web search, web fetch, advanced search, and scholarly paper search via hosted Exa and Parallel services without requiring an API key.
README
pi-search
Free, model-agnostic web search for pi. It uses the hosted Exa and Parallel MCP services, the same search strategy used by opencode. No API key is required.
Tools
| Tool | Use |
|---|---|
web_search |
Everyday web discovery with clean, answer-ready excerpts. |
web_fetch |
Read selected public URLs when search excerpts are insufficient. |
web_search_advanced |
Filtered Exa searches by category, domain, date, freshness, and context budget. |
search_papers |
Scholarly publications, preprints, and journal articles through Exa's publication index. |
Use semantic queries that describe the ideal page. For broad topics, web_search accepts
2-3 concise queries for Parallel's multi-query search. Search results are excerpt-first;
only call web_fetch for exact wording, a full-page analysis, or conflicting results.
Advanced search
web_search_advanced supports Exa's category indexes (news, company, people,
publication, personal site, and financial report), domain allow/deny lists, publication
date ranges, and maxAgeHours freshness. Set maxAgeHours: 0 to force a live crawl.
It defaults to eight results, 10,000 context characters per result, and token-efficient
highlights.
search_papers fixes the category to publication and defaults to eight results, 4,000
characters per paper, and highlights. It accepts date ranges, type: "deep" for complex
literature reviews, and enableSummary for triage. The publication index does not support
domain filters.
Provider selection
web_search and web_fetch select Exa or Parallel deterministically per pi session. Override
the basic-tool provider with PI_SEARCH_PROVIDER:
PI_SEARCH_PROVIDER=exa pi
PI_SEARCH_PROVIDER=parallel pi
web_search_advanced and search_papers always use Exa because Parallel has no equivalent
filtering or publication index.
Both services have a free tier. Optional EXA_API_KEY and PARALLEL_API_KEY raise rate limits.
A few parameters are backend-specific: web_search.queries and web_fetch.objective /
fullContent are honored by Parallel and ignored by Exa, which searches query alone and
caps fetch size with maxCharacters. Unused parameters are dropped without error.
Install
pi install npm:@blessium/pi-search
To try it without installing:
pi -e npm:@blessium/pi-search
For a local checkout:
pi install /path/to/pi-search
Test
npm test
Security
The tools send your query text, the current pi session ID, and the active model name to the
hosted Exa and Parallel services over HTTPS. The extension makes no filesystem writes, runs
no shell commands, and stores no credentials locally. Optional EXA_API_KEY and
PARALLEL_API_KEY values are transmitted to the respective service only.
License
MIT
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.