searchio

Keyless web search and page reading for AI agents, built to get past bot walls.

Connector by IOServicesLabs · GitHub account verified by the MCP Registry

IndexedApache-2.0Local (stdio)

Install

Works with Claude Code, Codex, VS Code, Cursor, Windsurf, Claude Desktop. Read the repository before you give any tool access to your data.

npx agentmarket install connector/ioserviceslabs-searchio

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

About

Keyless web search and page reading for AI agents, built to get past bot walls.

Reviews

Write a review

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this server can do and what it asks for, from its tool list and install config. Checked Oct 25, 2026.

Tools: 9 read · 1 write · 0 delete · 2 not yet classified
Secrets: None asked for
Install scripts: Not scanned yet
Runs as: Node.js package

Tools by what they do 12

Read from the source code; the running server may list different tools.

Read 9 · Look things up without changing anything.

find_items
read_page
read_page_rendered
read_url
search
search_items
search_local
stats
web_search

Write 1 · Create, change, send or run things.

submit_report

Not yet classified 2 · We could not tell from the name and description.

research
youtube_transcript

Tools (12)

Read from the source code; the running server may differ. Descriptions are the server’s own words; hidden characters are shown as [U+…].

find_items
Find purchasable listings for a product across marketplaces and return normalized records with price, seller, availability and rating. Offers for the same product are merged, so each row is one product with its alternative offers attached. Use for any 'what does X cost' or 'where can I buy X' question.
read_page
Fetch one URL and return its readable text as markdown. Use this only after a search result looks genuinely worth reading -- it costs a network round trip and a large slice of your context. Handles bot-protected pages automatically by escalating transports. If the result says the page is an empty JavaScript shell, use read_page_rendered on the same URL instead.
read_page_rendered
Fetch one URL through the full-browser fidelity tier (real Chromium: executes every kind of JavaScript including module scripts and WASM, carries real session state, solves managed challenges). This is 10-100x slower and heavier than read_page, so use it ONLY when read_page came back as an empty/JS-shell page or you already know the content only exists after scripts run. For ordinary pages read_page is always the right call.
read_url
Fetch one URL through the blocking-resistant acquisition ladder. Escalates plain HTTP -> TLS impersonation -> stealth browser as the target demands, and returns the page as extracted markdown. Args: url: Absolute URL to read.
research
Run the multi-step research swarm on a question (needs an LLM key). Decomposes the question, runs parallel research workers over the live web, and returns a cited answer with findings, items and sources. Args: question: The research question, min 3 chars.
search
Fused, ranked web search across keyless providers. Args: query: What to search for. intent: web | shopping | news | video ... k: Number of results (1-50). domains: Comma-separated allowlist, e.g. "example.com,other.org". freshness: all | day | week | month | year. locale: Language/market of the question, e.g. de-DE.
search_items
Marketplace/product listings with prices, merged across retailers. Args: query: Product to find, e.g. "sony wh-1000xm5". k: Maximum listings to return. domains: Comma-separated site scope, e.g. "bestbuy.com".
search_local
Facebook Marketplace classifieds near a city (browser-backed). Args: query: What to look for, e.g. "kayak". city: Marketplace city slug, e.g. "nyc". Empty follows the exit IP. k: Maximum listings.
stats
Ladder and provider health (per-domain learned tiers, block rates).
submit_report
Finish this subtask and return your findings. Call this exactly once, when you have enough evidence or have concluded the evidence is not available. Every claim must carry at least one citation URL you actually retrieved. If you could not answer, say so in gaps rather than guessing.
web_search
Search the web and get ranked results with titles and snippets. Results are fused across several independent providers, so a result found by more than one is more likely to be right. Use short keyword queries (2-6 words), not sentences. Run several narrow searches rather than one broad one.
youtube_transcript
Pull the caption transcript of one YouTube video. Uses the caption track YouTube already publishes -- no audio download, no transcription -- so it is fast and the transcript of record when it exists. Returns timed segments plus the full plain text. A video without captions, or a string that is not a video, is an error naming which; it is never an empty success. Args: url: watch/youtu.be/shorts URL or bare 11-char video id. lang: Preferred caption language, e.g. "en", "de".

Signals

Stars: 0
Forks: 0
Open issues: 0
Weekly downloads: 44
Last commit: Oct 10, 2026 (18 days ago)
Last release: Oct 11, 2026 (17 days ago)
Last package publish: Oct 11, 2026 (17 days ago)

From GitHub (fetched Oct 24, 2026) and the package registry.

Details

Registry name: io.github.IOServicesLabs/searchio
Version: 0.1.1
Auth: No auth
Packages: npm @ioserviceslabs/searchio-mcp
License: Apache-2.0
Repository: github.com/IOServicesLabs/searchio
Website: www.ioservices.io/searchio
First seen: Oct 24, 2026
Last checked: Oct 24, 2026
Source: Official MCP Registry github-graphql