--- title: "Uploadcare Integration" description: "Ingest files into Uploadcare from public URLs, manage the stored library with filters and pagination, inspect file and group metadata, and build on-the-fly CDN transformation URLs for resizing, cropping, and format conversion without an extra API round trip." url: https://flowrunner.ai/integrations/uploadcare date_modified: 2026-08-04T10:04:16-07:00 --- # Uploadcare [Storage](https://flowrunner.ai/integrations/category/storage) Ingest files into Uploadcare from public URLs, manage the stored library with filters and pagination, inspect file and group metadata, and build on-the-fly CDN transformation URLs for resizing, cropping, and format conversion without an extra API round trip. [Verified](https://flowrunner.ai/integrations/verified "What does verified mean?") · 10 actions · API key · available [Uploadcare website](https://uploadcare.com/) · [Platform Documentation](https://uploadcare.com/docs/) · Capability data verified 2026-08-12 1. A new product with supplier-hosted images lands in the catalog system 2. Upload File From URL fetches each image server-side into the Uploadcare project 3. Get Upload Status polls the token until the file reaches success and yields its UUID 4. Get File Info inspects dimensions, MIME type, and size against the storefront's media standards 5. Agent flags images below resolution thresholds or in unsupported formats 6. The merchandising lead reviews flagged images before they reach the live storefront, because a pixelated hero image is a brand decision no threshold fully captures 7. Store File persists approved images so the 24-hour temporary window never eats them 8. Build CDN URL generates the resized, WebP-optimized delivery variants for the product page 9. The product record updates with UUIDs and delivery URLs attached ## What This Integration Enables Uploadcare's sharpest idea is that a file and its variants are not the same thing: store one original, derive everything else in the delivery URL. This connector gives FlowRunner agents the full lifecycle around that idea. They ingest files from public URLs without touching local storage, poll asynchronous uploads to completion, manage the library with filters and pagination, persist or remove files deliberately, inspect metadata and groups, and compose transformation URLs, resize, crop, format conversion, quality, preview, with no API round trip at all. Media operations that used to be a queue of design requests become steps in a flow. - Ingest supplier, user, and system files server-side from public URLs - Poll upload tokens so downstream steps only run against ready files - Enforce media standards by inspecting dimensions and MIME types before publish - Persist production files explicitly and let ephemeral ones expire by design - Generate transformed delivery URLs per surface without storing variants ### Without FlowRunner **Images pass through laptops**: Someone downloads supplier files, resizes them in an editor, and re-uploads, per image **Temporary files vanish mysteriously**: Uploads that nobody stored disappear after a day, and the gap surfaces as a broken page **Every variant is a stored copy**: Thumbnails, banners, and previews multiply as physical files nobody tracks ### With FlowRunner **Ingestion is server-side and polled**: Files fetch from source URLs, and flows wait on upload status before depending on them **Persistence is explicit**: Store File runs as a deliberate step, so nothing production-bound rides the 24-hour window **Variants are URLs, not files**: Resize, crop, format, and quality are CDN operations composed on the fly ## Use Case Scenarios ### Product images from supplier link to storefront-ready A product syncs from [Shopify](https://flowrunner.ai/integrations/shopify) with a supplier image URL. The agent ingests it with Upload File From URL, polls Get Upload Status to success, checks Get File Info against the resolution floor, stores it, and calls Build CDN URL three times: "resize/1200x1200" with "format/webp" for the product page, "crop/400x400/center" for the grid, "preview" with "quality/smart" for search results. One stored original, three delivery URLs, zero image-editor sessions. ### The library audit that runs itself Monthly, the agent pages List Files with storage-state filters and writes UUID, filename, size, and stored state to [Google Sheets](https://flowrunner.ai/integrations/google-sheets). Temporary files still referenced by active flows get flagged for storing; stored files with no known reference land on a removal-candidate list for a human to review. Get Project Info confirms autostore settings and collaborators as part of the same sweep. Storage costs stay explained, and the library stays legible. ### Archive the original, deliver the optimized For compliance, originals must live in company-controlled storage. After ingestion completes, the agent passes the file's CDN URL to [AWS S3](https://flowrunner.ai/integrations/s3) for durable archival, then posts the confirmation to [Slack](https://flowrunner.ai/integrations/slack) with both locations. Uploadcare serves the fast, transformed variants; S3 holds the untouched original. Each system does what it is actually for, and the flow is the glue nobody has to maintain by hand. ## Human-in-Loop Highlight Delete File is irreversible, and its blast radius is invisible at the API: the UUID being deleted may be embedded in live product pages, email templates, and cached CDN URLs across the internet, and the moment datetime\_removed is set, every one of those references breaks at once. A cleanup agent acting on "not referenced in the systems I can see" is exactly the kind of confident-and-wrong that turns a tidy library into a storefront full of broken images. So FlowRunner cleanup flows separate discovery from destruction: the agent assembles removal candidates from List Files and Get File Info, upload date, size, stored state, whatever reference evidence exists, and presents the docket. "Delete 47 stored files, oldest 26 months, no references found in catalog or templates?" The media owner, who knows where else URLs may live, approves the list or trims it. Expiring temporaries handle themselves by design; deleting stored files is a signature event. Agent processes routinely Detects exception requiring judgment Clear match Continues automatically Ambiguous Routes to human via email Human decides Agent resumes with decision ## Agent Capabilities 10 actions ### Ingestion 2 - **Upload File From URL** Fetches a file server-side from a public URL into the project, optionally storing it permanently on upload. Returns a token for status polling. No local file handling required. - **Get Upload Status** Polls an upload token through waiting, progress, and success, where the response yields the file UUID and metadata, or an error status. The gate that keeps flows from building on files that are not ready. ### Library Management 4 - **List Files** Lists project files with pagination, filters for stored versus temporary and removed state, and ordering by upload time or size. The audit and cleanup sweep primitive. - **Get File Info** Retrieves one file's metadata by UUID: original filename, size, MIME type, image dimensions and content info, storage state, and CDN URLs. The inspection step behind media standards enforcement. - **Store File** Persists a file so the roughly 24-hour temporary cleanup never removes it. The explicit commitment that separates production files from ephemera. - **Delete File** Permanently removes a file by UUID, setting datetime\_removed. Irreversible, and gated behind a human docket in every FlowRunner cleanup flow. ### Groups 2 - **List Groups** Lists file groups, bundles of files under a single group ID, with pagination and creation-time ordering. - **Get Group Info** Retrieves a group by its uuid~count ID with file count, timestamps, group CDN URL, and the member file list. ### Project and Delivery 2 - **Get Project Info** Retrieves project name, public key, autostore setting, and collaborators. The credential check and configuration read. - **Build CDN URL** Composes a transformed delivery URL from a UUID and chained operations, resize/500x500, crop/400x400/center, format/webp, quality/smart, preview, with no API call made. The action that makes variants free. ## Frequently Asked Questions ### What can FlowRunner do with Uploadcare? FlowRunner agents can run Upload File From URL, Get Upload Status, and List Files in Uploadcare, plus 7 more actions. ### Does connecting Uploadcare to FlowRunner require OAuth? No. Uploadcare connects to FlowRunner with an API key, no OAuth flow required. ### Can Uploadcare trigger a FlowRunner workflow automatically? Uploadcare doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger. **Work at Uploadcare?** This integration exposes Uploadcare to AI agents on every FlowRunner plan, including through MCP, at no cost to you. [See what FlowRunner offers integration partners](https://flowrunner.ai/integrations/partners), including how to keep this page current. --- Markdown version of https://flowrunner.ai/integrations/uploadcare. Site index: https://flowrunner.ai/llms.txt