Skip to main content
Glama
DeepSeek Search Everywhere MCP

DeepSeek Search Everywhere MCP

deepseek_extract_many

Read-onlyIdempotent

Extract the same fields or JSON schema from up to 10 web pages in parallel, returning structured data with verbatim evidence quotes per URL in input order.

Instructions

Pull the same fields or JSON schema out of up to 10 web pages in parallel, with evidence quotes.

$0.00075 per URL plus LLM tokens; input across pages defaults to 60,000 chars (max_chars_per_url 20,000). Returns data and evidence per URL in input order; values without a verbatim quote are null and a failed URL is reported in place. Output is sanitized.

When NOT to use: one URL (deepseek_extract), plain page text (deepseek_fetch_many), finding pages (deepseek_search_many), or a topic report (deepseek_research).

Returns: {"ok": True, "results": [{"url": str, "ok": bool, "data": {...}, "evidence": {...}}], "partial": bool, "cost_usd": float}.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "urls": "<urls>"
    },
    "name": "deepseek_extract_many"
  }
}

Input Schema

NameRequiredDescriptionDefault
urlsYesList of HTTP/HTTPS URLs to extract structured data from concurrently
cleanNoRemove Markdown backslash escapes from page text before extraction (default True)
fieldsNoList of field names to extract (mutually exclusive with schema)
schemaNoJSON schema describing output structure (mutually exclusive with fields)
refreshNoBypass local cache and fetch fresh pages from upstream (default False)
deadlineNoOverall timeout deadline in seconds; partial results returned on expiry
instructionNoOptional natural-language extraction guidance (default None)
max_total_charsNoMax total characters across all pages for extraction (default 60000)
max_chars_per_urlNoMax characters per page to feed to extraction LLM (default 20000)
strip_boilerplateNoRemove cookie banners and boilerplate before extraction (default True)

Output Schema

NameRequiredDescriptionDefault

No arguments

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed10 schema fields changedv3.0.0
    • addedInput schema / properties / clean / description
      Added value: +"Remove Markdown backslash escapes from page text before extraction (default True)"
    • addedInput schema / properties / deadline / description
      Added value: +"Overall timeout deadline in seconds; partial results returned on expiry"
    • addedInput schema / properties / fields / description
      Added value: +"List of field names to extract (mutually exclusive with schema)"
    • addedInput schema / properties / instruction / description
      Added value: +"Optional natural-language extraction guidance (default None)"
    • addedInput schema / properties / max_chars_per_url / description
      Added value: +"Max characters per page to feed to extraction LLM (default 20000)"
    • addedInput schema / properties / max_total_chars / description
      Added value: +"Max total characters across all pages for extraction (default 60000)"
    • addedInput schema / properties / refresh / description
      Added value: +"Bypass local cache and fetch fresh pages from upstream (default False)"
    • addedInput schema / properties / schema / description
      Added value: +"JSON schema describing output structure (mutually exclusive with fields)"
    • addedInput schema / properties / strip_boilerplate / description
      Added value: +"Remove cookie banners and boilerplate before extraction (default True)"
    • addedInput schema / properties / urls / description
      Added value: +"List of HTTP/HTTPS URLs to extract structured data from concurrently"
  2. Addedv2.2.0

TDQS

A4.4/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Beyond the annotations (readOnly, idempotent, openWorld), the description discloses per-URL pricing plus LLM tokens, result ordering matching input order, null-on-no-verbatim-quote semantics, in-place reporting of failed URLs, partial-result behavior, and output sanitization. These are real operational facts an agent could not learn from annotations or schema. Minor gap: it doesn't state a concurrency or rate limit.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loaded with purpose, then a clean when-NOT list and a returns block; each section is clearly demarcated. Slightly redundant is the returns line, since an output schema already exists and the schema already states the char defaults restated in the opening paragraph.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 10-parameter, cost-incurring batch tool with rich annotations and an output schema, the description covers everything else that matters: pricing, ordering guarantees, null/failure semantics, partial results, and sibling routing. Nothing an agent needs before calling is missing.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so every parameter is already documented; the description's mention of the 60,000-char total and 20,000-per-URL defaults only restates the schema. The 'fields or JSON schema' phrasing hints at the mutual exclusivity the schema already declares, adding little new meaning.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource — pull the same fields or JSON schema out of up to 10 web pages in parallel — plus a distinctive output trait (evidence quotes). The batching scope (up to 10 URLs) immediately separates it from the single-URL sibling.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Explicitly enumerates four when-NOT cases and names the correct alternative for each: one URL, plain text, page discovery, topic report. This is the strongest form of routing guidance, leaving no inference required.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.