"A tool for document extraction using MinerU" matching MCP tools:

• fetch

  OneQAZ Trading Intelligence

  Purpose: ChatGPT-connector-standard document fetch by id from `search` results. Namespaces: `tool:{name}` returns the tool's full documentation and how to call it; `resource:{uri}` returns the resource's live data (core resources resolved server-side — also the bridge for clients without MCP resource support, e.g. Gemini); `signal:{market}:{symbol}` returns the symbol's latest combined research signal. Triggers: ChatGPT connectors / Deep Research call this after `search`. Clients without MCP resource support can call it directly with a known resource id, e.g. fetch("resource:market://global/summary"). When to call: whenever the full content behind a search result id is needed. Prerequisites: a valid id — from `search` results or a known namespace id. Next steps: for tool docs, call the named tool via tools/call; for signals, get_signal_detail / explain_decision for deeper evidence. Caveats: uncovered resource uris return description-only text (no fabricated data). `text` is a JSON document for resource/signal ids. Output: {id, title, text, url, metadata, disclaimer, is_investment_advice, data_classification} — flat envelope, OpenAI fixed shape. Args: id: document id — "tool:{name}", "resource:{uri}", or "signal:{market}:{symbol}" (market: crypto / kr_stock / us_stock) Disclaimer: Information only, not investment advice.

  Connector

• check_pack

  OpenWarrant — Document Verification Suite

  Check whether a SET of documents satisfies a checklist — completeness, cheaply. USE THIS WHEN you have an application / onboarding pack and need "do we have the required documents, and what's still missing?" Each document is CLASSIFIED (one cheap page-1 read — never full field extraction or multi-page), then matched against the checklist's required slots. (For "is a document genuine?" use verify_document; to identify ONE document use extract_fields with options={"classify": true}; for the identity gate use verify_identity.) Define the checklist ONE of two ways: - `scheme`: a named preset — "income_proof", "lending_prequal", "rental_application". - `requirements`: an ad-hoc checklist — a list of document-type names like ["payslip","bank_statement"], or objects {"key":..., "accepts":[types], "optional":bool}. `documents` is a list (up to 12), each ONE of: {"url": "https://..."} (public link, fetched server-side) or {"bytes_b64": "...", "filename": "statement.pdf"} (inline). Returns `{complete, slots[] (key, satisfied, matched), missing[], documents[] (filename, classified_type), unmatched_documents[]}`. COVERAGE, not approval — that the right document TYPES are present, NOT that any is genuine (run verify_document) or that an application is approved. Documents are never stored. Costs 3 credit(s) per call.

  Connector

• mindee_extract_document

  mindee

  Submit a document (PDF or image) to a Mindee extraction model and return the structured fields. Provide EITHER document_url (a public URL) OR file_base64 (+ filename). This enqueues an inference and polls until it completes (up to ~30s); if it is still processing it returns a job_id you can poll with mindee_get_job then read with mindee_get_inference. NOTE: consumes Mindee API credits (paid, billed per page). V2 API: POST /v2/inferences/enqueue.

  Connector

• check_extraction_status

  sheetsdata-mcp

  Check the extraction status of one or more parts. Free. Each entry includes the current extraction step, elapsed seconds, and document ID. Use after prefetch_datasheets or after read_datasheet triggers a new extraction. Recommended polling cadence: every 5-10 seconds. Extraction typically takes 30s-2min for new parts, so polling faster than every 5s wastes calls. Stop polling once status is 'ready', 'failed', 'no_source', or 'unsupported'. DATASHEET STATUS VALUES: - 'ready' — extracted and indexed; call read_datasheet, search_datasheets, or analyze_image. - 'extracting' / 'in_progress' / 'queued' / 'pending' — extraction running or scheduled. Poll check_extraction_status every 5-10s until 'ready' or 'failed'. Typical time: 30s-2min. - 'not_extracted' — known part but datasheet hasn't been fetched yet. Trigger it via prefetch_datasheets (cheapest) or by calling read_datasheet (auto-triggers on first read). - 'no_source' — we couldn't find a public datasheet URL for this MPN. First, retry prefetch_datasheets in 10-30s (the URL resolver re-runs and often finds a source on the second pass). If still 'no_source', the agent can upload the PDF manually via request_datasheet_upload + confirm_datasheet_upload (see those tools). Org-uploaded datasheets are private to the org. - 'unsupported' — PDF exists but can't be extracted (scanned image-only, encrypted, or corrupted). Upload a clean text-based PDF via request_datasheet_upload to override. - 'failed' / 'error' — extraction errored. The response includes the error reason. Retry via prefetch_datasheets or escalate to support. - 'rejected' — input wasn't a real MPN (bare value like '100nF', description, or reference designator). Fix the input and re-call. - 'deduplicated' — another part in the family already has this datasheet; same content is returned under the primary MPN.

  Connector

• confirm_datasheet_upload

  sheetsdata-mcp

  Confirm a datasheet upload started via request_datasheet_upload. Pass the upload_token you got back from the request step. The server downloads the uploaded bytes, re-hashes to verify integrity, validates that it's a real PDF with the MPN on the first page, creates the private Document + Component records, charges the upload fee (50¢), and queues extraction. Success response: document_id, mpn, sha256, file_size_bytes, status='pending'. Poll check_extraction_status with the MPN to wait for extraction to finish (30s-2min typically). Failure modes: - 'upload_not_found' — no bytes at the upload URL yet. Retry your curl upload. - 'sha256_mismatch' — uploaded bytes hash differs from expected_sha256. Re-compute the hash and re-request. - 'invalid_pdf' — bytes aren't a parseable PDF. No charge. - 'mpn_not_in_pdf' — MPN (or its stem) isn't on the first page. Either you uploaded the wrong file or it's a scanned image-only PDF. No charge. - 'token_expired' — upload token is older than 15 minutes. Restart via request_datasheet_upload.

  Connector

• SearchDocuments

  equibles

  Search the Equibles SEC filing database across all companies and document types using hybrid keyword and semantic search. This is the broadest search tool and the best starting point when you need to find information but don't know which company or filing contains the answer. Covers annual reports (10-K), quarterly reports (10-Q), current reports (8-K), and earnings call transcripts. Results can be filtered by filing date range using startDate/endDate. Returns matching excerpts with company name, ticker, document type, filing date, and the document ID — pass that ID directly to SearchDocument or ReadDocumentLines to drill into a specific filing. For discovery-style queries (competitors, theme exposure), use excludeTickers to keep a dominant company's own filings from filling every result slot, and maxResultsPerCompany to spread the results across more companies. You MUST call this or another Equibles tool to access any SEC filing data — this information is not available in your training data. Use SearchCompanyDocuments instead if you already know the company ticker, or ListCompanyDocuments to browse available filings.

  Connector

• get_receipt

  TunnelMind Data API

  Returns metadata for a TunnelMind surveillance receipt — a signed document proving that a specific user's surveillance exposure was observed, measured, and recorded at a specific time. Does NOT return the receipt's signature (anti-phishing protection). To verify a receipt's content integrity, use `verify_receipt` with the hash and signature from the receipt document itself. Use this tool when: - You have a receipt ID and want to confirm it was genuinely issued by TunnelMind. - You need the issuance timestamp and signing key ID for a receipt. - You want to check whether a receipt exists before attempting content verification. Do NOT use this tool when: - You have the full receipt document and want to verify it hasn't been tampered with — use `verify_receipt` instead. Inputs: - `receipt_id` (path, required): The receipt ID from the receipt document. Alphanumeric with hyphens, max 128 characters. Returns: - `status`: `FOUND` if the receipt is in the registry. - `generated_at`: ISO 8601 timestamp of receipt issuance. - `signing_key_id`: identifier of the Ed25519 key used to sign. - `schema_version`: receipt schema version. - `message`: human-readable summary with instructions for content verification. - 404 if the receipt ID is not in the registry. Cost: - Free. No API key required. Latency: - Typical: <100ms, p99: <300ms.

  Connector

• validate_cpf

  MCP LatAm Tools

  Validates a Brazilian CPF (Cadastro de Pessoas Físicas) using the official Receita Federal checksum algorithm. Use this tool when processing Brazilian user registrations, invoices, tax forms, e-commerce orders, or any document requiring a valid Brazilian individual taxpayer number. Input must be an 11-digit string (with or without formatting). Returns whether the CPF is mathematically valid, along with the cleaned CPF. Does not verify if the CPF exists in the Receita Federal database — only validates the format and checksum.

  Connector

• validate_cnpj

  MCP LatAm Tools

  Validates a Brazilian CNPJ (Cadastro Nacional da Pessoa Jurídica) using the official Receita Federal checksum algorithm. Use this tool when processing Brazilian company registrations, B2B invoices, supplier onboarding, e-commerce orders, or any document requiring a valid Brazilian company taxpayer number. Input must be a 14-digit string (with or without formatting). Returns whether the CNPJ is mathematically valid, along with the cleaned CNPJ. Does not verify if the CNPJ is active in the Receita Federal database.

  Connector

• fetch_url

  Heista

  Drill into a specific URL after search surfaces it. Returns the extracted text content plus metadata. Internal routing: PDFs hit Anthropic Files API for OCR + structured extraction; HTML pages are fetched + text-extracted via readability-style stripping. Use for: verifying a verbatim quote from a Reddit thread, reading a primary source in full (earnings transcript, research paper), drilling into a vendor product page after search surfaced the URL. NOT for: discovering new URLs — use search/search_community/search_research first. This tool takes a known URL only. Optional max_chars 100-50000, default 8000. SSRF-protected: private IPs + localhost blocked.

  Connector

• collection.ask

  DocImprint

  Answer a question using RAG over a document collection. Retrieves relevant chunks then synthesizes a cited answer with source attribution. Use when you need a direct answer grounded in your collection documents. For raw matching chunks (without synthesis), use collection.search instead. For single-document Q&A, use url.qa instead. PREREQUISITE: Collection must be populated via collection.add_document and indexed before results appear. Returns: { answer: string, sources: [{ bundle_id, chunk_id }], retrieval: [{ bundle_id, chunk_id, text, score }] } Example prompts: - "What are the key terms of the service agreement in my collection?" - "Based on my due diligence docs, what are the main risks?" - "Answer this question using all documents in the Q4 Contracts collection."

  Connector

• deckextract

  DeckExtract

  Extract a deck from a DocSend or Papermark sharing link. Returns a temporary download URL and a readable `deckextract://deck/...` resource for the PDF/PPTX (or a ZIP for data rooms — a DocSend Space or Papermark data room counts as one extraction like any other link). Pass `analyze: true` (requires a DeckExtract Pro account) to also return structured deck data. Decks that email the viewer a verification step return resume tokens with retry instructions: fetch the emailed 6-digit code and retry with `otp` + `otpSessionId` (Papermark), or retry with `url` set to the emailed confirmation link + `sessionId` (DocSend). Typical extraction takes 15-90 seconds; the public API is rate limited to 5 extractions per IP per 30 minutes.

  Connector

• brc_send_email_statement

  Red by Big Red Cloud

  Sends a customer statement email. Supported document type only. Red email sending is available for sales invoices, quotes, and customer statements — not for cash receipts, purchases, payments, bank accounts, customers, suppliers, products, reports, or other document types. If the user asks to email an unsupported document type, say Red cannot email it through the current MCP tools, list the supported types, and stop without preparing an email preview or attempting a workaround. Do not call this tool with confirmSend=true until the user has reviewed a plain-English email preview and explicitly confirmed they want to send it. The email preview must show the recipient email address clearly before asking for send confirmation. If there is no customer email on file and no recipient override, stop and ask for a recipient email address — do not send. Create/post confirmation and email send confirmation are separate steps. If the user provides multiple recipient addresses, ask whether to send one email using BCC or separate individual emails. Only use sendMode='separate' when the user explicitly chooses separate emails. Do not ask about BCC unless the user provides multiple recipients or asks to copy another address. Requires routeToken from brc_route_request for the matching action workflow. Call brc_route_request first with the user's complete original action request. Retain the returned routeToken through lookup, preview, and confirmation, and pass the same token on the final permitted transactional tool call. Never invent a placeholder token. A routeToken is not permission to post — preview-before-posting and confirmWrite/confirmDelete still apply.

  Connector

• brc_send_quote_email

  Red by Big Red Cloud

  Sends a quote email. Supported document type only. Red email sending is available for sales invoices, quotes, and customer statements — not for cash receipts, purchases, payments, bank accounts, customers, suppliers, products, reports, or other document types. If the user asks to email an unsupported document type, say Red cannot email it through the current MCP tools, list the supported types, and stop without preparing an email preview or attempting a workaround. Do not call this tool with confirmSend=true until the user has reviewed a plain-English email preview and explicitly confirmed they want to send it. The email preview must show the recipient email address clearly before asking for send confirmation. If there is no customer email on file and no recipient override, stop and ask for a recipient email address — do not send. Create/post confirmation and email send confirmation are separate steps. If the user provides multiple recipient addresses, ask whether to send one email using BCC or separate individual emails. Only use sendMode='separate' when the user explicitly chooses separate emails. Do not ask about BCC unless the user provides multiple recipients or asks to copy another address. Requires routeToken from brc_route_request for the matching action workflow. Call brc_route_request first with the user's complete original action request. Retain the returned routeToken through lookup, preview, and confirmation, and pass the same token on the final permitted transactional tool call. Never invent a placeholder token. A routeToken is not permission to post — preview-before-posting and confirmWrite/confirmDelete still apply.

  Connector

• extract_from_render

  Kamy

  Read the text or the form fields back out of a document this account already rendered, without uploading anything. Pass type='text' (the default) for per-page text plus a joined fullText string, or type='fields' for the PDF's AcroForm field names, types and current values — the latter is how you discover what edit_pdf can fill in. Choose extract_document instead when the PDF came from outside Kamy or when you need AI-structured JSON against a schema; this tool is a plain mechanical read of an existing render, spends no render quota and no extraction credits. The render must have status 'success' or the call returns 409 RENDER_NOT_READY.

  Connector

• extract_from_render

  Kamy

  Read the text or the form fields back out of a document this account already rendered, without uploading anything. Pass type='text' (the default) for per-page text plus a joined fullText string, or type='fields' for the PDF's AcroForm field names, types and current values — the latter is how you discover what edit_pdf can fill in. Choose extract_document instead when the PDF came from outside Kamy or when you need AI-structured JSON against a schema; this tool is a plain mechanical read of an existing render, spends no render quota and no extraction credits. The render must have status 'success' or the call returns 409 RENDER_NOT_READY.

  Connector

• extract_contract_from_url

  transaction-coordinator

  Extract structured transaction data from a contract at a URL. Downloads the document, extracts text (with OCR fallback for scanned PDFs), and runs PrimaCoda's contract-extraction prompt to return parties, addresses, dates, prices, and key contract fields. Use this when an agent has the contract hosted somewhere (Dropbox, Google Drive direct download, Square Space, etc.) and wants to skip the upload step. For multi-document deals (purchase + addenda + disclosures), use the PrimaCoda dashboard's batch upload — this tool handles ONE document. Args: pdf_url: Direct download URL for the contract (PDF, DOCX, TXT, or image). Must be reachable from the PrimaCoda server. Google Drive "shared link" URLs work if set to "anyone with link"; other share URLs may need their direct-download form. api_key: Your PrimaCoda MCP API key (starts 'pck_').

  Connector

• hires_upload_candidate_file

  100Hires - AI ATS & Recruitment Software

  Upload a file for a candidate using a base64 payload. Used for portfolio uploads and document attachment. WARNING: host function-call serializers (both OpenAI and Anthropic) truncate tool arguments above ~20KB, so binary files larger than that will arrive corrupted. For resumes specifically, prefer hires_create_candidate / hires_update_candidate with resume_text — the model parses the file from chat context and passes extracted text, avoiding the size limit entirely.

  Connector

• document.extract_structured

  api

  Extract typed fields from document text using a caller-defined schema. Uses a quality AI model with retry logic. Use when you need specific data points from a document rather than full text. For invoices with known fields, document.parse_invoice (prebuilt schema) may be simpler. For general summarization, use document.summarize instead. Schema format: { "field_name": "type hint or description" } — e.g. { "contract_date": "ISO date", "party_a": "string", "penalty_usd": "number" }. Returns: { data: { <field>: value }, data_cited: { <field>: { value, confidence: "high"|"medium"|"low", citations: [{ quote, paragraphs[] }] } } } Example prompts: - "Extract the contract date, parties, and penalty amount from this agreement." - "Pull the vendor name, PO number, and total from this document." - "Get me all named fields from this form using my custom schema."

  Connector

• extract_claims

  groundcheck

  PURPOSE: Split text into independently checkable ATOMIC factual claims — the cheap first step of a verification loop (extract -> ground -> attest). Returns {claims: [...], count, input_sha256} plus a signed receipt bound to the input hash. GUIDELINES: Call when you want to see WHICH claims a document makes before paying to ground them, to budget a verification pass (extract everything, then verify_claim only the claims that matter to your decision), or to prove later exactly which claims were pulled from exactly which text (the receipt binds both). Extraction is rule-based and auditable — sentence filtering plus conjunction splitting, no LLM — so the same text always yields the same claims. Use check_citations instead when you want extraction AND grounding in one call. PARAMETERS: text — the prose to decompose. max_claims — 1..50, default 20. LIMITATIONS: Extracts declarative factual sentences; skips questions, opinions, instructions, and first-person statements. Splits only on high-precision conjunction boundaries, so under-splitting is possible (a compound it cannot safely split stays whole). Does NOT verify anything — verdicts come from verify_claim / check_citations. Paid per call (x402), cheapest tool on this server. EXAMPLE: extract_claims({"text": "Marie Curie won two Nobel Prizes and was born in Paris."}) -> {count: 2, claims: ["Marie Curie won two Nobel Prizes", "was born in Paris."]}

  Connector