Introducing News: 200M articles indexed live

28 PDF & File Parsing Prompts for ChatGPT & Claude

Long PDFs, scanned forms, and messy spreadsheets are where AI assistants lose detail. These file parsing prompts send the document through Context’s Parse API, which converts PDF, Word, Excel, PowerPoint, and image files to clean Markdown with OCR and page markers, so the assistant works from the full text and can cite the page.

File Parsing prompts

  • Contract key terms

    Parse this contract: [file]. Extract parties, term, renewal, termination rights, payment terms, liability caps, indemnities, and governing law into a table with page references. Flag anything unusual.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Invoices to JSON

    Parse these invoices: [files]. Return JSON per invoice with vendor, invoice number, date, due date, line items with quantity and unit price, tax, and total. Check that the line items add up to the total.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Bank statement categorizer

    Parse this bank statement: [file]. List every transaction with date, description, amount, and a spending category, then total each category and flag recurring charges.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • 10-K risk summary

    Parse [company]’s latest 10-K: [file or URL]. Summarize the business, segment revenue, top ten risk factors, and anything new compared with the prior year, citing page numbers.

    Uses: Parse Files, Scrape, Web Search. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Resumes to structured profiles

    Parse these resumes: [files]. Return JSON with name, contact, roles with dates, skills, education, and certifications for each. Then rank them against this job description: [JD].

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • RFP requirements matrix

    Parse this RFP: [file]. Build a requirements matrix with each requirement, its section and page, whether it’s mandatory, and a column for our response. List deadlines and submission rules separately.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Research paper explainer

    Parse this paper: [file]. Explain the question, method, data, main results, and limitations in plain English, and list the three figures or tables that matter most with page numbers.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Scanned forms to data

    Use OCR to parse these scanned forms: [files]. Extract [fields] into a CSV and flag fields that were illegible or empty instead of guessing.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Spreadsheet sanity check

    Parse this spreadsheet: [file]. Describe each sheet, find errors, blanks, duplicates, and outliers, and summarize what the data says in five bullets.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Pitch deck to memo

    Parse this pitch deck: [file]. Turn it into a one-page investment memo covering problem, solution, market, traction, team, and ask, plus the five questions I should ask the founders.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Insurance policy comparison

    Parse these insurance policies: [files]. Compare coverage limits, deductibles, exclusions, waiting periods, and premiums side by side, and point out gaps with page references.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Lease abstract

    Parse this commercial lease: [file]. Produce a lease abstract with the rent schedule, escalations, term, options, maintenance obligations, assignment rules, and critical dates.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Compare two versions of a document

    Parse [version 1] and [version 2] of this document. List every substantive change, added clause, and removed clause, and ignore formatting differences.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Medical bill check

    Parse this medical bill and explanation of benefits: [files]. Match each charge to what insurance paid, flag duplicates or charges that don’t match the visit, and draft questions for the billing office.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Financial statements to tables

    Parse these financial statements: [file]. Extract the income statement, balance sheet, and cash flow statement into clean tables, then calculate margins, growth, and current ratio for each year.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Handbook Q&A with citations

    Parse our employee handbook: [file]. Answer these questions with the exact policy text and page number: [questions]. Say “not covered” when the handbook doesn’t address one.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Receipts to expense report

    Parse these receipt photos with OCR: [files]. Extract merchant, date, amount, tax, and category, and format them as an expense report. Flag any line over [threshold].

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Spec sheet comparison

    Parse these product spec sheets: [files]. Build a comparison table of [specs] across all products, keep the units stated on each sheet, and mark where a spec is missing.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Action items from meeting minutes

    Parse these meeting minutes: [files]. Extract every decision, action item, owner, and deadline into a table, and list items that were raised but never resolved.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Government PDF from a link

    Fetch and parse the PDF at [URL]. Summarize what it requires, who it applies to, deadlines, and penalties, citing page numbers.

    Uses: Scrape, Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Translate and summarize a document

    Parse this [language] document: [file]. Translate the key sections into English, summarize the whole document, and keep the original wording for legal or technical terms.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Syllabi to one calendar

    Parse these syllabi: [files]. Extract every assignment, exam, and reading with its due date into a single calendar table sorted by date.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Annual report KPIs across years

    Parse [company]’s annual reports for the last [N] years: [files or URLs]. Extract [KPIs] for each year into one table and describe the trend.

    Uses: Parse Files, Scrape. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Data room review

    Parse the files in this data room: [files]. Build an index of documents, summarize each in two lines, and list documents a buyer would expect for [deal type] that are missing.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Code and config explainer

    Parse these code or config files: [files]. Explain what each one does and how they connect, then list risky or outdated settings in a table with file, setting, risk, and suggested fix.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • PDF tables to clean CSVs

    Parse [file] and extract every table into its own CSV with the page number it came from. Repeat merged-cell values so each row stands alone, move units into the headers, and mark unreadable cells as “unclear” instead of guessing.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Files to RAG-ready chunks

    Parse these files: [files]. Convert each to Markdown that keeps headings, lists, and tables, then split it at headings into chunks of about [500] tokens as JSONL with file name, page number, section path, and text.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT
  • Customer interview synthesis

    Parse these interview transcripts: [files]. For each, pull the pains, current workarounds, and buying triggers with short verbatim quotes, and keep hypotheticals like “I would use it” separate. Then build one table showing how many interviews raised each theme.

    Uses: Parse Files. Install the Context.dev MCP if you haven’t already.

    Open inClaudeChatGPT

How to use these prompts

  1. Install the Context.dev MCP

    Add the Context.dev MCP server to Claude, Cursor, or another agent, or install the ChatGPT app. This gives the assistant live web search, scraping, and data tools.

  2. Pick a prompt

    Copy it, or open it straight in Claude or ChatGPT. Every copied prompt tells the assistant to use Context.dev, links the MCP install page, and asks it to cite sources.

  3. Fill in the brackets

    Replace every [bracket] with your details, such as a company domain, a competitor, or a file, then run it.

Context APIs behind these prompts

The tools your assistant calls when it runs these prompts.

File Parsing prompt FAQ

Which file types do these prompts support?

PDF, DOCX, DOC, XLSX, XLS, PPTX, CSV, HTML, Markdown, JSON, RTF, and images such as PNG, JPG, and TIFF. Scanned PDFs and images can be read with OCR.

Can the assistant cite page numbers?

Yes. Parse can mark the start of each PDF page in its output, so prompts here ask for a page reference next to every extracted fact.

Can it read a PDF from a link instead of an upload?

Yes. Scrape can fetch and parse a PDF at a public URL, including a page range, so you can point the assistant at a report link instead of downloading it first.

How do I install the Context.dev MCP in Claude or ChatGPT?

Add the Context.dev MCP server to Claude, Cursor, or any MCP client with the MCP setup guide, or install the Context app in ChatGPT for web search and page reading. Then paste a prompt and the assistant picks the right Context tools.

Do these prompts work without Context?

They run, but without live web tools the assistant answers from its training data and can’t open the pages, files, or searches the prompt asks for. Expect older information and more guesses.

More prompt categories

Start building with live web context.

Scrape, search, and research the web with one API. Bring the results straight into your product.

1,000free credits

Every month. No credit card required.

Start for freeRead the docs