> ## Documentation Index > Fetch the complete documentation index at: https://hc.starbridge.ai/llms.txt > Use this file to discover all available pages before exploring further. # Get Buyer File Download Links > Returns 1-hour signed download URLs for a single buyer file: the original file (in its native format — PDF, DOCX, VTT, etc.) and the parsed-markdown rendering. Use `fileId` from `researchBuyerFiles` (per-file IDs in `buyerFiles[].id`). Prefer `viewFileContents` for token-bounded markdown; use this tool only when you need the raw bytes (e.g., to render a PDF for the user, hand off to a parser, or attach the original to an email). **WARNING — both files may be very large.** Original PDFs are commonly multi-megabyte; parsed markdown is usually smaller but neither size is bounded. Do NOT fetch either link and inline its contents into your context. For text, use `viewFileContents` (token-bounded). To show a file to the user, hand them the URL directly. Response fields: - `originalLink` — 1h signed URL to the original file. May be very large; see the warning above. - `markdownLink` — 1h signed URL to the parsed-markdown rendering, or `null` if the file is not yet processed (e.g., status != Processed). May also be very large; see the warning above. - `originalSizeBytes` — size of the original blob in bytes. May be null for older files. Use this to decide whether to fetch. - `contentType` — MIME type of the original (e.g., `application/pdf`, `text/vtt`). - `type` — buyer-file classification (`Rfp`, `Contract`, `PurchaseOrder`, `BoardMeeting`, `StrategicPlan`, `Other`, …); `null` if not yet detected. - `originalName` — original file name as uploaded or scraped. For scraped sources this is often a URL-encoded source path (e.g., `https_www_youtube_com_watch_v_xyz.vtt`) — do NOT surface it to the user as-is. Links expire in 1 hour. Returns 404 if no buyer file with the given id exists. ## OpenAPI ````yaml https://core-http2-157602306522.us-central1.run.app/external/public/swagger/documentation.yaml get /api/external/buyer/file/{fileId}/download-links openapi: 3.1.1 info: title: Starbridge External Public API version: '1.0' servers: - url: https://dashboard.starbridge.ai security: [] paths: /api/external/buyer/file/{fileId}/download-links: get: tags: - External MCP - External API - Buyer summary: Get Buyer File Download Links description: >- Returns 1-hour signed download URLs for a single buyer file: the original file (in its native format — PDF, DOCX, VTT, etc.) and the parsed-markdown rendering. Use `fileId` from `researchBuyerFiles` (per-file IDs in `buyerFiles[].id`). Prefer `viewFileContents` for token-bounded markdown; use this tool only when you need the raw bytes (e.g., to render a PDF for the user, hand off to a parser, or attach the original to an email). **WARNING — both files may be very large.** Original PDFs are commonly multi-megabyte; parsed markdown is usually smaller but neither size is bounded. Do NOT fetch either link and inline its contents into your context. For text, use `viewFileContents` (token-bounded). To show a file to the user, hand them the URL directly. Response fields: - `originalLink` — 1h signed URL to the original file. May be very large; see the warning above. - `markdownLink` — 1h signed URL to the parsed-markdown rendering, or `null` if the file is not yet processed (e.g., status != Processed). May also be very large; see the warning above. - `originalSizeBytes` — size of the original blob in bytes. May be null for older files. Use this to decide whether to fetch. - `contentType` — MIME type of the original (e.g., `application/pdf`, `text/vtt`). - `type` — buyer-file classification (`Rfp`, `Contract`, `PurchaseOrder`, `BoardMeeting`, `StrategicPlan`, `Other`, …); `null` if not yet detected. - `originalName` — original file name as uploaded or scraped. For scraped sources this is often a URL-encoded source path (e.g., `https_www_youtube_com_watch_v_xyz.vtt`) — do NOT surface it to the user as-is. Links expire in 1 hour. Returns 404 if no buyer file with the given id exists. operationId: getFileDownloadLinks parameters: - name: fileId in: path description: Buyer file ID, from `researchBuyerFiles`. required: true schema: type: string format: uuid responses: '200': description: Signed download links and metadata for the requested buyer file. content: application/json: schema: $ref: '#/components/schemas/BuyerFileDownloadLinks' '400': description: Invalid request body or format '401': description: Missing or invalid authentication '403': description: Unauthorized to access resource or perform action '404': description: Resource not found '415': description: Unsupported media type or request body format '429': description: Rate limit exceeded '500': description: Internal server error security: - apiKey: [] components: schemas: BuyerFileDownloadLinks: type: object title: BuyerFileDownloadLinks required: - id - contentType - originalLink properties: id: type: string description: Unique identifier of the buyer file. format: uuid originalName: type: - string - 'null' description: >- Original file name as uploaded or scraped. For scraped sources this is often a URL-encoded source path (e.g., `https_www_youtube_com_watch_v_xyz.vtt`) — do NOT surface it to the user as-is. contentType: type: string description: >- MIME type of the original file (e.g., `application/pdf`, `text/vtt`). type: type: - string - 'null' description: >- Buyer-file classification (Rfp, Contract, PurchaseOrder, BoardMeeting, StrategicPlan, Other, …). Null if the type has not been detected yet. enum: - Rfp - RfpProposal - Contract - PurchaseOrder - VendorList - BoardMeeting - BoardMeetingAgenda - BoardMeetingAgendaPacket - StrategicPlan - CapitalImprovementPlan - AnnualFinancialReport - ProposedBudget - Budget - BudgetAmendment - CurriculumAdoptionSchedule - Other originalSizeBytes: type: - integer - 'null' description: >- Size of the original file in bytes. May be null for older files where size was not recorded. Use this to decide whether the file is too large to fetch. originalLink: type: string description: >- 1-hour signed URL to the original file. WARNING: this file may be very large (multi-megabyte PDFs are common) and is not size-bounded. Do NOT fetch and inline its contents into your context — use `viewFileContents` for token-bounded text, or hand the URL to the user. markdownLink: type: - string - 'null' description: >- 1-hour signed URL to the parsed-markdown rendering of the file, or null if the file is not yet processed. WARNING: markdown is typically smaller than the original but is also not size-bounded and may itself be very large. Do NOT fetch and inline its contents into your context — use `viewFileContents` for token-bounded text, or hand the URL to the user. securitySchemes: apiKey: scheme: bearer description: HTTP Bearer Authentication type: http ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.