Skip to content

PreteWorks API MCP server

com.preteworks/preteworks-apipreteworks.comstreamable-httpv1.0.2Scanned

One key for the document-to-web pipeline: scrape, AI-extract, build & edit PDFs, fill forms.

29 tools: 1 do something that can't be undone, 21 change things and 7 only read.

Tools: 0 delete, 1 irreversible, 21 write, 7 read.
  • Reads private data · 1
  • Reads text others write · 10
  • Can send data out · 10
  • 3 findings

Tools listed without sign-in. Remote posture

With Anlyon

1 tool here does something that can't be undone. Govern it.

  • Anlyon lists the tools and adopts none.
  • Each tool you adopt gets approvals and receipts.
  • Your agent never holds the key.
Govern this server →

A tool your agent calls directly is not routed through Anlyon and is not governed by it. Anlyon Radar only lists and labels, and these labels are the same either way.

What the PreteWorks API MCP server can do

The PreteWorks API MCP server is published by preteworks.com, a domain the MCP Registry verified. It has 29 tools. It can do things that cannot be undone, such as sending, paying or publishing (merge_pdfs) and create or change data (21 tools: render_html_to_pdf, number_pdf, set_pdf_metadata, render_document and 17 more), but cannot delete data. 7 tools only read.

1 tool reads private data (summarize_document); 10 tools bring in text other people can write (batch_scrape, crawl_site, map_site and 7 more); 10 tools can send data out of your system (batch_scrape, crawl_site, map_site and 7 more). On its own it has all three parts of a prompt-injection exfiltration path: private data, text others write, and a way out.

The fixed checks matched 3 findings (annotation-conflict). Each one comes with the text it matched.

Its remote endpoint lists its tools without sign-in.

Last scanned Nov 14, 2026, version 1.0.2.

Tools

Grouped by what each tool does, most consequential first. Open one for its reasons and full description; descriptions are the publisher's own words.

29 tools

Can't be undone 1Claude Code: ask
  • merge_pdfsCombine 2+ PDFs into a single PDF, in the order given.No findings

    Why it is in this group

    • The name "merge_pdfs" says "merge": a word for a change that another call does not undo.
    Description, as the server wrote it
    Combine 2+ PDFs into a single PDF, in the order given. Inputs are file_ids (from prior tools) or base64 PDFs. Returns a file_id (for chaining) and a ~1h download URL.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • openWorldHint: false
Changes things 21Claude Code: ask
  • answer_from_documentAnswer a question grounded ONLY in a source document — a web page, PDF, Office file, or raw text.Reads text others writeCan send data outNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Answer a question grounded ONLY in a source document — a web page, PDF, Office file, or raw text. Provide 'question' plus ONE source: 'url', 'pdf', 'file' (base64), or 'text'. Says so when the answer isn't in the document.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: true
  • data_to_spreadsheetTurn rows of data into a downloadable XLSX (default) or CSV.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Turn rows of data into a downloadable XLSX (default) or CSV. 'rows' is an array of objects (keys → header row) or an array of arrays (first row is the header). Returns a file_id (for chaining) and a ~1h download URL.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • openWorldHint: false
  • docx_to_pdfConvert a .docx document (base64) to a PDF.No findings

    Why it is in this group

    • The description's first verb is "convert": a word for a change.
    Description, as the server wrote it
    Convert a .docx document (base64) to a PDF. Returns a file_id and a ~1h URL.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • openWorldHint: false
  • docx_to_textExtract the text of a .docx document (base64).No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Extract the text of a .docx document (base64). Returns the text inline.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: false
  • extract_dataExtract specified fields from a web page, PDF, or Office file as JSON, using AI.Reads text others writeCan send data outNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Extract specified fields from a web page, PDF, or Office file as JSON, using AI. Provide 'fields' plus ONE source: a 'url', a 'pdf' (file_id or base64), or a 'file' (base64 .docx/.xlsx/.csv). Returns a JSON object mapping each field to its value, or null when absent.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: true
  • extract_pdf_textExtract the text content of a PDF — for RAG, summarization, or search.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Extract the text content of a PDF — for RAG, summarization, or search. Accepts a file_id (from a prior tool) or a base64-encoded PDF, and returns the text inline. Not OCR: a scanned/image-only PDF returns little or no text.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: false
  • file_to_markdownConvert a Word (.docx), Excel (.xlsx) or CSV file (base64) into clean Markdown — for RAG, agents and pipelines.1 finding

    Why it is in this group

    • The description's first verb is "convert": a word for a change.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    Findings

    • Warn

      The server marks "file_to_markdown" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

      Matched text

      annotations: {"readOnlyHint":true,"openWorldHint":false}; The description's first verb is "convert": a word for a change.

      What this check looks for
    Description, as the server wrote it
    Convert a Word (.docx), Excel (.xlsx) or CSV file (base64) into clean Markdown — for RAG, agents and pipelines. DOCX keeps headings/lists/tables; spreadsheets become Markdown tables (one per sheet). Returns Markdown inline. Optional 'format' (docx|xlsx|csv) overrides auto-detection.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: false
  • fill_pdf_formFill a PDF's form fields from a map of field -> value; optionally flatten.No findings

    Why it is in this group

    • The name "fill_pdf_form" says "fill": a word for a change.
    Description, as the server wrote it
    Fill a PDF's form fields from a map of field -> value; optionally flatten. Returns a file_id and a ~1h URL.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • openWorldHint: false
  • images_to_pdfCombine PNG/JPEG images (base64) into a PDF, one image per page.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Combine PNG/JPEG images (base64) into a PDF, one image per page. Returns a file_id and a ~1h URL.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • openWorldHint: false
  • map_siteDiscover the same-origin URLs linked from a page — the site map, with no page content fetched.Reads text others writeCan send data outNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Discover the same-origin URLs linked from a page — the site map, with no page content fetched. Private/internal hosts are blocked.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: true
  • markdown_to_pdfRender Markdown to a clean, print-styled PDF.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Render Markdown to a clean, print-styled PDF. Returns a file_id and a ~1h URL.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • openWorldHint: false
  • number_pdfStamp a page number on every page.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Stamp a page number on every page. position: bottom-center (default) | bottom-left | bottom-right | top-center | top-left | top-right. format supports {n} and {total}. Returns a file_id + ~1h URL.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • openWorldHint: false
  • render_documentRender a structured document to a PDF from a named template (invoice, receipt, report).No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Render a structured document to a PDF from a named template (invoice, receipt, report). Money/totals are computed for you. Returns a file_id (for chaining) and a ~1h download URL.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • openWorldHint: false
  • render_html_to_pdfRender a full HTML document to a PDF.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Render a full HTML document to a PDF. External subresources are blocked for safety — inline images/fonts as data: URIs. Returns a file_id (for chaining) and a ~1h download URL.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • openWorldHint: false
  • rotate_pdfRotate every page of a PDF by 90, 180, or 270 degrees (clockwise).No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Rotate every page of a PDF by 90, 180, or 270 degrees (clockwise). Returns a file_id (for chaining) and a ~1h download URL.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • openWorldHint: false
  • select_pagesKeep only the specified pages of a PDF (e.g.No findings

    Why it is in this group

    • The name "select_pages" says "select": a word for a change.
    Description, as the server wrote it
    Keep only the specified pages of a PDF (e.g. [1,3,5]) and drop the rest, preserving order. Returns a file_id (for chaining) and a ~1h download URL.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • openWorldHint: false
  • set_pdf_metadataSet a PDF's title, author, subject, keywords (array or comma string) and/or creator.No findings

    Why it is in this group

    • The name "set_pdf_metadata" says "set": a word for a change.
    Description, as the server wrote it
    Set a PDF's title, author, subject, keywords (array or comma string) and/or creator. Only the fields you provide change. Returns a file_id + ~1h URL.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • openWorldHint: false
  • split_pdfSplit a PDF into multiple PDFs — one per page by default, or by page ranges like '1-3;4-6'.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Split a PDF into multiple PDFs — one per page by default, or by page ranges like '1-3;4-6'. Returns a file_id + URL per part.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • openWorldHint: false
  • url_to_pdfFetch a public http/https URL and render the live page to a PDF.Reads text others writeCan send data out1 finding

    Why it is in this group

    • The description's first verb is "fetch": a word for reading.
    • The server marks it as not read-only (readOnlyHint: false).

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe name "url_to_pdf" says "url": it reads web pages, which anyone can write.
    • untrusted-inputThe description's first sentence says "fetch" and "http": it reads web pages, which anyone can write.
    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Findings

    • Warn

      The server marks "url_to_pdf" as not read-only (readOnlyHint: false), and its words say it reads.

      Matched text

      annotations: {"readOnlyHint":false,"destructiveHint":false,"openWorldHint":true}; The description's first verb is "fetch": a word for reading.

      What this check looks for
    Description, as the server wrote it
    Fetch a public http/https URL and render the live page to a PDF. Returns a file_id (chainable into the PDF tools) and a ~1h download URL. Private/internal hosts are blocked.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • openWorldHint: true
  • url_to_screenshotFetch a public http/https URL and capture a PNG screenshot.Reads text others writeCan send data out1 finding

    Why it is in this group

    • The name "url_to_screenshot" says "screenshot": a word for reading.
    • The server marks it as not read-only (readOnlyHint: false).

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe name "url_to_screenshot" says "screenshot" and "url": it reads web pages, which anyone can write.
    • untrusted-inputThe description's first sentence says "fetch" and "http": it reads web pages, which anyone can write.
    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Findings

    • Warn

      The server marks "url_to_screenshot" as not read-only (readOnlyHint: false), and its words say it reads.

      Matched text

      annotations: {"readOnlyHint":false,"destructiveHint":false,"openWorldHint":true}; The name "url_to_screenshot" says "screenshot": a word for reading.

      What this check looks for
    Description, as the server wrote it
    Fetch a public http/https URL and capture a PNG screenshot. Set full_page for the entire scroll height. Returns a file_id and a ~1h download URL.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • openWorldHint: true
  • watermark_pdfStamp a diagonal grey text watermark (e.g.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Stamp a diagonal grey text watermark (e.g. "DRAFT", "CONFIDENTIAL") across every page of a PDF. Returns a file_id (for chaining) and a ~1h download URL.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • openWorldHint: false
Only reads 7Show 7 read-only tools
  • batch_scrapeFetch up to 10 public URLs and return each as clean Markdown, in one call — for research/RAG over several pages at once.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The description's first verb is "fetch": a word for reading.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe name "batch_scrape" says "scrape": it reads web pages, which anyone can write.
    • untrusted-inputThe input takes a URL (input.urls), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.urls): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Fetch up to 10 public URLs and return each as clean Markdown, in one call — for research/RAG over several pages at once. Private/internal hosts are blocked.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: true
  • crawl_siteFetch a URL plus up to 7 more same-origin pages it links to (≤8 total), each as clean Markdown.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The description's first verb is "fetch": a word for reading.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe name "crawl_site" says "crawl": it reads web pages, which anyone can write.
    • untrusted-inputThe description's first sentence says "fetch" and "url": it reads web pages, which anyone can write.
    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Fetch a URL plus up to 7 more same-origin pages it links to (≤8 total), each as clean Markdown. Bounded and synchronous. Private/internal hosts are blocked.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: true
  • pdf_infoRead a PDF's metadata (title, author, subject, keywords, creator, producer, dates) plus page count and page size, as JSON.No findings

    Why it is in this group

    • The description's first verb is "read": a word for reading.
    Description, as the server wrote it
    Read a PDF's metadata (title, author, subject, keywords, creator, producer, dates) plus page count and page size, as JSON.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: false
  • read_pdf_formList a PDF's form fields (name, type, value, options) as JSON.No findings

    Why it is in this group

    • The name "read_pdf_form" says "read": a word for reading.
    Description, as the server wrote it
    List a PDF's form fields (name, type, value, options) as JSON.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: false
  • read_urlFetch a public http/https URL and return its main content as clean Markdown — ideal for giving an agent readable web content for research or RAG.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "read_url" says "read": a word for reading.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe name "read_url" says "read" and "url": it reads web pages, which anyone can write.
    • untrusted-inputThe description's first sentence says "fetch" and "http": it reads web pages, which anyone can write.
    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Fetch a public http/https URL and return its main content as clean Markdown — ideal for giving an agent readable web content for research or RAG. Private/internal hosts are blocked.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: true
  • scrape_pageFetch a public http/https URL and return its rendered HTML, the links on the page, and page metadata (title/description/OpenGraph/canonical/favicon) — in one render.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The description's first verb is "fetch": a word for reading.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe name "scrape_page" says "scrape": it reads web pages, which anyone can write.
    • untrusted-inputThe description's first sentence says "fetch" and "http": it reads web pages, which anyone can write.
    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Fetch a public http/https URL and return its rendered HTML, the links on the page, and page metadata (title/description/OpenGraph/canonical/favicon) — in one render. For clean Markdown use read_url instead. Private/internal hosts are blocked.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: true
  • summarize_documentSummarize a web page, PDF, Office file (.docx/.xlsx/.csv), or raw text using AI.Reads private dataReads text others writeCan send data outNo findings

    Why it is in this group

    • The name "summarize_document" says "summarize": a word for reading.

    What it can reach

    • private-dataThe name "summarize_document" says "summarize" and "document": it reads documents.
    • private-dataThe description's first sentence says "summarize" and "file": it reads files.
    • private-dataThe input names a file (input.file), and the tool reads: it reads files.
    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe description's first sentence says "summarize" and "web": it reads web pages, which anyone can write.
    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Summarize a web page, PDF, Office file (.docx/.xlsx/.csv), or raw text using AI. Provide ONE source: 'url', 'pdf' (file_id or base64), 'file' (base64), or 'text'. Optional 'max_words'. Returns the summary.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: true

This is not a safety rating, and no finding is not a safety guarantee. The groups and findings come from fixed checks of the tool definitions the server returned on the date shown, not from its code or what it does when called. How the checks work.

Exfil surface

Which tools read private data, which read text someone else can write, and which can send data out. One of each is a path a prompt injection can use.

100 paths in this server alone from private data, through text someone else wrote, to a way out. Denying 1 private-data tool breaks every one.

Reads private data1
Reads files, mail, messages, databases, secrets, calendars or repositories.
Reads text others write10
Brings in text someone else can write: web pages, inbound email, messages, issues and comments.
Can send data out10
Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.
deny to break every path
{
  "permissions": {
    "deny": [
      "mcp__preteworks-api__summarize_document"
    ]
  }
}

Other servers in the same agent add their own tools to these paths. Audit your whole setup to see them together.

Remote posture

What anyone on the internet sees at the remote endpoint: an anonymous initialize and tools/list, and the public OAuth metadata.

Tools listed without sign-in

Looked at Nov 20, 2026, 18:45 UTC, unchanged since Oct 22, 2026. No credentials were sent and no tool was called. What is looked at.

Reachable
yes
HTTP status of an anonymous initialize
200
tools/list served without credentials
yes
Tools that change things, listed without credentials
22
Protocol version
2025-06-18
WWW-Authenticate on 401
not observed
resource_metadata in WWW-Authenticate
not observed
RFC 9728 protected resource metadata
not observed
Authorization servers
not observed
Legacy HTTP+SSE transport
no
Dynamic client registration
not observed
Client ID metadata documents
not observed
  • HighThe whole server

    The endpoint served tools/list without credentials, and the list has 22 tools that change things (1 irreversible, 21 write). Whether a call also works without credentials was not tested: tools are never called.

    Matched text

    https://api.preteworks.com/mcp: tools/list returned 29 tools to an anonymous client, including merge_pdfs (irreversible), render_html_to_pdf (write), number_pdf (write), set_pdf_metadata (write), render_document (write) and 17 more

    What this check looks for

Findings

Matches of fixed checks, with the text they matched.

3 warn
  • The server marks "file_to_markdown" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

    Matched text

    annotations: {"readOnlyHint":true,"openWorldHint":false}; The description's first verb is "convert": a word for a change.

    What this check looks for
  • The server marks "url_to_pdf" as not read-only (readOnlyHint: false), and its words say it reads.

    Matched text

    annotations: {"readOnlyHint":false,"destructiveHint":false,"openWorldHint":true}; The description's first verb is "fetch": a word for reading.

    What this check looks for
  • The server marks "url_to_screenshot" as not read-only (readOnlyHint: false), and its words say it reads.

    Matched text

    annotations: {"readOnlyHint":false,"destructiveHint":false,"openWorldHint":true}; The name "url_to_screenshot" says "screenshot": a word for reading.

    What this check looks for

History

Changes to its tool definitions and to its remote endpoint, newest first.

No change seen yet. Its tools are rescanned every day, and a remote endpoint is looked at once a week.

Install

Pinned to version 1.0.2. Values in angle brackets are yours to fill in.

claude mcp add --transport http preteworks-api https://api.preteworks.com/mcp --header 'Authorization: <your-authorization>'
claude mcp add --transport http preteworks-api https://api.preteworks.com/mcp --header 'Authorization: <your-authorization>'

Run in a terminal.

Claude Code permissions

Allow the reads, ask before a change or a way out, deny the deletes. Paste into .claude/settings.json.

.claude/settings.json
{
  "permissions": {
    "allow": [
      "mcp__preteworks-api__pdf_info",
      "mcp__preteworks-api__read_pdf_form"
    ],
    "ask": [
      "mcp__preteworks-api__render_html_to_pdf",
      "mcp__preteworks-api__number_pdf",
      "mcp__preteworks-api__set_pdf_metadata",
      "mcp__preteworks-api__render_document",
      "mcp__preteworks-api__merge_pdfs",
      "mcp__preteworks-api__select_pages",
      "mcp__preteworks-api__rotate_pdf",
      "mcp__preteworks-api__watermark_pdf",
      "mcp__preteworks-api__extract_pdf_text",
      "mcp__preteworks-api__images_to_pdf",
      "mcp__preteworks-api__markdown_to_pdf",
      "mcp__preteworks-api__split_pdf",
      "mcp__preteworks-api__fill_pdf_form",
      "mcp__preteworks-api__file_to_markdown",
      "mcp__preteworks-api__data_to_spreadsheet",
      "mcp__preteworks-api__batch_scrape",
      "mcp__preteworks-api__crawl_site",
      "mcp__preteworks-api__map_site",
      "mcp__preteworks-api__extract_data",
      "mcp__preteworks-api__summarize_document",
      "mcp__preteworks-api__answer_from_document",
      "mcp__preteworks-api__docx_to_text",
      "mcp__preteworks-api__docx_to_pdf",
      "mcp__preteworks-api__read_url",
      "mcp__preteworks-api__url_to_pdf",
      "mcp__preteworks-api__url_to_screenshot",
      "mcp__preteworks-api__scrape_page"
    ],
    "deny": []
  }
}

Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. A read that can send data out (it takes a URL, say) asks first too. It assumes the server is installed as preteworks-api. Review it before you use it.

Capability tags behind 10 rules
  • mcp__preteworks-api__batch_scrape untrusted-input, outbound
  • mcp__preteworks-api__crawl_site untrusted-input, outbound
  • mcp__preteworks-api__map_site untrusted-input, outbound
  • mcp__preteworks-api__extract_data untrusted-input, outbound
  • mcp__preteworks-api__summarize_document private-data, untrusted-input, outbound
  • mcp__preteworks-api__answer_from_document untrusted-input, outbound
  • mcp__preteworks-api__read_url untrusted-input, outbound
  • mcp__preteworks-api__url_to_pdf untrusted-input, outbound
  • mcp__preteworks-api__url_to_screenshot untrusted-input, outbound
  • mcp__preteworks-api__scrape_page untrusted-input, outbound

Pin

Keep the decision: fail CI when the tools change, follow every change, or show a badge.

Pin in CI

Fail a build when the tools you reviewed are not the tools the server serves today.

curl -s "https://radar.anlyon.com/api/v1/servers/com.preteworks/preteworks-api/verify?hash=c41dfc636fa077dc3e570bf91149a6f1c4f7fff12eab0907c62d272c46ba2d0a" | jq -e '.changed == false'

Pinned hash c41dfc636fa0, from Nov 14, 2026, 02:18 UTC.

Watch

Get told when its tools change: a new tool, a new description, a new class.

https://radar.anlyon.com/servers/com.preteworks/preteworks-api/changes.xml

An Atom feed. Works in any feed reader or Slack's RSS app.

Badge

Anlyon Radar badge for com.preteworks/preteworks-api
[![Anlyon Radar](https://radar.anlyon.com/badge/com.preteworks/preteworks-api.svg)](https://radar.anlyon.com/servers/com.preteworks/preteworks-api)

Source: MCP Registry. Synced Oct 21, 2026, 18:00 UTC.

PreteWorks API MCP server: questions and answers

Answered from the scan above, not from the publisher's description.

Is the PreteWorks API MCP server safe?

No scan can promise that. What its definitions show: 29 tools, 22 of which can change, send or delete something, and 3 findings from fixed checks. Allow the reads, ask before each change, and deny the deletes; the permissions block on this page does that.

Can the PreteWorks API MCP server delete data?

No tool is classed as destructive: going by their definitions, none of them deletes, cancels or revokes anything.

Can the PreteWorks API MCP server do things that cannot be undone?

Yes. 1 tool can send, pay, publish or merge: merge_pdfs.

What tools does the PreteWorks API MCP server have?

29 tools: 1 irreversible, 21 write, 7 read. merge_pdfs, render_html_to_pdf, number_pdf, set_pdf_metadata, render_document, select_pages, rotate_pdf, watermark_pdf and 21 more.

Does the PreteWorks API MCP server need sign-in or OAuth?

Its remote endpoint listed its tools to an anonymous client, without sign-in.

Does the PreteWorks API MCP server send data to outside services?

10 tools can carry data out (it sends, posts or publishes, or takes a recipient, channel or URL): batch_scrape, crawl_site, map_site, extract_data, summarize_document, answer_from_document and 4 more.

Has the PreteWorks API MCP server changed its tools?

Not since Anlyon Radar started tracking it. Last scanned Nov 14, 2026.