Spy Search follows a versioned roadmap that is developed openly with the community. The current stable release is v0.3, shipped on 2025-06-10. Work is ongoing toward v0.4 and v0.5.Documentation Index
Fetch the complete documentation index at: https://mintlify.com/JasonHonKL/spy-search/llms.txt
Use this file to discover all available pages before exploring further.
v0.3 — Released (2025-06-10)
All planned features for v0.3 have shipped:- User Interface — React-based frontend for interacting with the search engine
- PDF search support — Query and extract content from PDF documents
- REST API — Programmatic access to Spy Search functionality
- Log output — Added application logging
- Local document search (RAG) — Retrieval-Augmented Generation over your own files
- Report generation with PDF export — Produce long-form, sourced reports (~2,000 words) and export them as PDFs
- Docker deployment — Official
docker-composesupport for easy self-hosting
v0.4 — Planned
Email, Notion & Obsidian API integrations
Email, Notion & Obsidian API integrations
Connect Spy Search to your existing productivity tools. Pull content from email, Notion workspaces, and Obsidian vaults as first-class search sources.
MCP (Model Context Protocol) support
MCP (Model Context Protocol) support
Add support for the Model Context Protocol, enabling Spy Search to act as an MCP-compatible tool server that other LLM clients can invoke directly.
Google Search API integration
Google Search API integration
Replace or supplement the current DuckDuckGo backend with the Google Search API for broader and more reliable web coverage.
Multi-site search
Multi-site search
Extend search beyond the current default provider to support querying multiple search engines and custom site targets in a single run.
Webhook support
Webhook support
Trigger searches and receive results via webhooks, enabling event-driven integrations with external systems and automation pipelines.
Markdown HTTP request handling
Markdown HTTP request handling
Support structured Markdown-formatted HTTP requests as an input format, making it easier to script and chain agentic search queries.
Refactoring pass
Refactoring pass
A dedicated internal refactor to improve code quality, maintainability, and developer experience ahead of the v0.5 performance work.
v0.5 — Planned
Fast Search (Compiled Language Module)
The current Python-based search path prioritises flexibility and report quality over raw speed. v0.5 will introduce a high-performance search tier written in a compiled language to dramatically reduce query latency. An experimental Go-based search server already exists in the repository (see Go web search module below) and is being used for early performance testing.Audio Summary Agent
A new agent that converts search reports into spoken audio summaries, making it possible to consume Spy Search results hands-free.Go Web Search Module
The repository includes an experimental Go-based HTTP search server in thewebsearch/ directory (websearch/main.go). It exposes a /search endpoint that accepts a q query parameter, fetches and parses result pages concurrently using goquery, and returns structured JSON results that include the URL, title, snippet, and per-URL elapsed timing. This module is the foundation for the compiled-language fast-search track planned in v0.5.
The roadmap is community-driven. If a feature matters to you, upvote or comment on the relevant GitHub Issue — or open a new one — to help the maintainers prioritise what ships next.