--- title: "pixx.io Integration" description: "pixx.io is a German digital asset management platform for images, videos, and documents. Agents push approved assets into the library with keywords and folders set, pull resized or converted versions, retag whole selections, and hand out shares and upload links." url: https://flowrunner.ai/integrations/pixx-io date_modified: 2026-09-04T14:07:24-07:00 --- # pixx.io [Storage](https://flowrunner.ai/integrations/category/storage) pixx.io is a German digital asset management platform for images, videos, and documents. Agents push approved assets into the library with keywords and folders set, pull resized or converted versions, retag whole selections, and hand out shares and upload links. 28 actions · API key · available [pixx.io website](https://pixx.io/) · [Platform Documentation](https://api.pixxio.com/docs/openapi) · Capability data verified 2026-08-13 1. A campaign asset is marked approved in the production tracker 2. Agent collects the asset file, its campaign, and the usage terms recorded against it 3. List Keywords reads the vocabulary already in use, so the tagging matches the rest of the library 4. Upload File pushes the asset in with its keywords, description, folder, and state set as it arrives 5. Create Collection groups the campaign's approved assets without moving them out of their folders 6. Agent posts the collection and the available download formats to the campaign channel 7. Before a door is opened for an outside agency to upload into the library, a person decides what state arriving files land in ## What This Integration Enables A digital asset manager earns its place when the library becomes the answer to "is this the approved version". pixx.io is built for that job, and three parts of its API say so. Every file carries a workflow state, so draft, approved, and archived are properties of the asset rather than a folder naming convention. Download formats are stored recipes, so the web crop, the print resolution, and the social ratio come out of the library instead of out of somebody's editor. And keywords are a finite shared vocabulary, which is the only thing that keeps a growing library findable. The connector exposes all three, which is what lets an agent participate in the library's discipline instead of quietly eroding it. FlowRunner agents push finished work in with its metadata already correct, pull the right rendition back out on request, retag a whole selection at once without disturbing the keywords already there, and hand out shares and upload links with an expiry. The [human-in-the-loop](https://flowrunner.ai/concepts/human-in-the-loop) gate on this page sits on the operation that decides what an outsider's files become the moment they arrive, because an asset that enters the library wearing the wrong state is worse than an asset that never arrived. ### Without FlowRunner **Assets scattered across drives**: The approved logo lives in a chat thread, a shared drive, and somebody's desktop, in three different crops **Resizing handled by request**: Every "can I get this at 1200 wide" is a person opening an editor and exporting by hand **Keywords invented per person**: The same subject gets tagged three ways, so none of the three finds all of it ### With FlowRunner **One library with a state on every asset**: Files carry draft, approved, or archived, so a flow can act on what is actually cleared **Resizing handled by recipe**: Download File takes a stored download format, and the right rendition comes out of the library **Keywords drawn from a shared vocabulary**: List Keywords is read before anything is tagged, so the library stays searchable as it grows ## Use Case Scenarios ### Approved work lands in the library already described A design is finalized in [Figma](https://flowrunner.ai/integrations/figma) and marked approved in the production tracker. The agent reads the campaign and subject from the tracker record, calls List Keywords so it tags with terms the library already knows, calls List File States so the asset lands in the state this workflow is allowed to set, and calls Upload File with the folder, description, keywords, and state supplied together. The asset is searchable from the first second it exists. Create Collection then gathers the campaign's approved pieces into one grouping while every file stays in its own folder, and the collection link goes to the campaign channel in [Slack](https://flowrunner.ai/integrations/slack). ### The right size, without a person in the middle A product listing needs imagery. The agent calls List Download Formats to see what renditions the library actually maintains, picks the one the channel needs, and calls Download File naming that format, so pixx.io does the conversion rather than a flow doing it badly. The rendition is written to the product record in [Shopify](https://flowrunner.ai/integrations/shopify) or the page in [Webflow](https://flowrunner.ai/integrations/webflow). When a partner asks for a set of images at once, Download Files As Zip bundles them in a single call. The library, not a person's export settings, decides what a web image looks like. ### Retagging a selection without breaking it A brand adopts a new product name. The agent calls List Files with a search term to gather everything tagged with the old name, then Update Files across the whole selection. Because Update Files adds and removes keywords rather than replacing the set, the campaign, photographer, and rights keywords already on those assets survive the change. Get File is called on a sample first so the agent can show the before and after of one asset to the brand owner, who is looking at a diff rather than at a promise, and Get Job reports when the run finishes. ## Human-in-Loop Highlight Create Upload Link is the operation that lets people outside the organization put files into the library, with keywords and a state applied as they arrive. That last part is the whole risk. Set the state to approved and everything an agency drops through that link becomes an approved brand asset the moment it lands, indistinguishable from work that a human actually cleared, and every downstream flow that filters on approved will pick it up and publish it. Nothing about that is undone by deleting the link afterwards. So when a flow needs an inbound door, the agent builds it and stops: "Ravenhill Studio needs to deliver the autumn shoot. I have the folder /Campaigns/Autumn/Raw and keywords autumn-2026, ravenhill, unretouched ready to apply. Which state should arrivals land in? Draft keeps them out of the published feed until somebody reviews them." The agent has read List File States and can name the options. It does not choose which one an outside contributor's unreviewed files get to wear. Agent processes routinely Detects exception requiring judgment Clear match Continues automatically Ambiguous Routes to human via preferred channel Human decides Agent resumes with decision ## Agent Capabilities 28 actions ### Files 7 - **List Files** Lists the files in the mediaspace, optionally narrowed by a search term or a folder. Returns one page at a time. - **Get File** Reads one file's metadata: its name, dimensions, keywords, folder, and the state it is in. - **Upload File** Uploads a file into the mediaspace from a URL your flow can reach, with its keywords, description, and folder set as it arrives. This is how an asset gets into pixx.io from anywhere else. - **Update Files** Changes several files at once. Keywords are added and removed rather than replaced, so a whole selection can be retagged without disturbing what is already there. - **Delete Files** Removes files from the mediaspace. - **Download File** Downloads a file into flow file storage and returns a link to it. A download format can be named to get a resized or converted version rather than the original. - **Download Files As Zip** Downloads several files as one ZIP into flow file storage. ### Folders 5 - **List Folders** Lists the folders in the mediaspace. - **Get Folder Tree** Returns the folder hierarchy rather than a flat list, so a flow can see what sits inside what. - **Create Folder** Creates a folder inside another one. pixx.io has no implicit root here, so a parent is always required. - **Rename Folder** Changes a folder's name. What is in it is untouched. - **Delete Folder** Removes a folder from the mediaspace. ### Collections 5 - **List Collections** Lists the collections in the mediaspace. A collection groups files without moving them out of their folders. - **Get Collection** Reads one collection and what is in it. - **Create Collection** Creates a collection, optionally with files already in it. Files stay where they are; a collection is a grouping, not a folder. - **Update Collection** Changes a collection's name, description, or contents. - **Delete Collections** Removes collections. The files in them stay in the mediaspace. ### External Shares 3 - **List External Shares** Lists the external shares on the mediaspace. An external share is a link that lets people outside your organization see or download chosen files. - **Create External Share** Creates a link that lets people outside your organization see chosen files. Giving email addresses emails them the link, so leave that empty to deliver it yourself. - **Delete External Shares** Removes external shares, which stops their links working. ### Upload Links 3 - **Create Upload Link** Creates a link that lets people outside your organization upload files into a chosen folder, with keywords and a state applied as they arrive. This page's human gate sits on the state. - **List Upload Links** Lists the upload links on the mediaspace and where each of them delivers. - **Delete Upload Links** Removes upload links, which stops people uploading through them. ### Library Vocabulary 3 - **List Keywords** Lists the keywords used across the mediaspace, which is what a flow needs to tag consistently rather than inventing new terms. - **List Download Formats** Lists the download formats set up on the mediaspace. Each one is a stored recipe for converting a file on the way out, and Download File takes one. - **List File States** Lists the workflow states a file can be in, such as draft, approved, or archived. Their identifiers are what the upload and update operations take. ### Account and Jobs 2 - **Get Current User** Reads the user the API key belongs to. Useful as a connection check, and worth reading because the key inherits that user's permissions. - **Get Job** Reads how a long-running operation is getting on. Operations run with the asynchronous option return a job rather than waiting. ## Frequently Asked Questions ### What can FlowRunner do with pixx.io? FlowRunner agents can run List Files, Get File, and Upload File in pixx.io, plus 25 more actions. ### Does connecting pixx.io to FlowRunner require OAuth? No. pixx.io connects to FlowRunner with an API key, no OAuth flow required. ### Can pixx.io trigger a FlowRunner workflow automatically? pixx.io doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger. **Work at pixx.io?** This integration exposes pixx.io 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/pixx-io. Site index: https://flowrunner.ai/llms.txt