
Export media from your application's owner-filtered records, then retrieve only the authorized completed tasks. The current Imagild API provides individual task status and content routes; it does not expose a “list every user's media” or account-wide bulk-export endpoint. Your application owns the export manifest and its delivery permissions.
Key Takeaways
- Select export items using the authenticated user's actual resource permissions.
- Keep task references, source files and generated results distinct.
- A failed download does not justify regenerating an already completed task.
Define the export scope before retrieving files
Let the user choose the relevant project, date range or media items through your application. Resolve those choices on the server; do not trust a browser-supplied list of raw remote task IDs as proof of ownership.
The Imagild API guide documents authenticated per-task access. If your backend shares one service account across application users, that account boundary does not replace your own user or workspace permission checks.
OWASP's authorization guidance recommends checking permissions on each request. Apply that principle to both building an export and downloading the completed archive.
Build a manifest with reviewable states
The following is an application-owned export record, not an Imagild API payload:
{
"exportId": "app-export-20261028-17",
"ownerScope": "AUTHORIZED_APPLICATION_USER",
"items": [
{
"applicationItemId": "poster-final",
"taskId": "SAVED_IMAGE_TASK_ID",
"kind": "image",
"exportState": "waiting_for_status",
"savedFilename": "poster-final.bin"
}
]
}
Keep the task identity, requested model and actual downloaded format with each result. Use generated or validated filenames; do not let an original upload name choose an arbitrary filesystem path inside an archive.
Avoid including API keys, session cookies, internal server configuration or every prompt automatically. If the user wants a prompt history, treat it as a deliberate export field subject to the same ownership filtering.
Retrieve completed media gradually
Check each permitted task's state. A pending task is not a missing file to recreate. A successful task can proceed to its authenticated content route, while a failed or unknown outcome should be represented accurately in the manifest.
Active API membership and a valid key remain required for reads and downloads. A 403 during an export is not evidence that the file was deleted. Inspect access before changing the media workflow.
Pace task reads and downloads within the account's current shared 120-read-per-minute budget. Large exports should use bounded concurrency and streaming or controlled temporary storage rather than loading all media into one unbounded memory buffer.
Validate the bytes and protect the archive
Check response status, content type and decoded media before marking an item exported. The HTML-instead-of-image guide explains why a saved error page must not pass as an image.
If the archive is delivered through your application, require its owner's authorization again. Your chosen expiry, storage destination and sharing behavior are your application's policies, not built-in properties of an Imagild export endpoint.
Keep a copy of the manifest with partial failures so the user can distinguish omitted, pending and successfully saved items. The browser download behavior guide helps separate retrieving media from presenting a reliable save action.
Test cross-user and partial-failure cases
With local fixtures, give two application users separate items under one simulated remote account. An export for one user must exclude the other's records even if their IDs appear in the requested filter.
Then simulate one valid result, one pending task and one failed download. The export should retain the valid result and a clear partial manifest without generating replacements or claiming all files were saved.
Pass the feature when ownership filtering, byte validation and archive delivery agree. These checks describe a proposed application export flow; they do not establish a privacy certification, deletion guarantee or new administrative API permission.
Related tasks
- Why Does My Downloaded Image Open as a Web Page?
- Why Does My Download Button Open the Photo in a New Browser Page Instead of Saving It?
Sources and scope
- Imagild API console: Actual model API authentication, membership, task ownership, supported routes and private media retrieval.
- OWASP Authorization Cheat Sheet: Validate resource permissions on each request, including access to static or exported resources.
Product instructions were checked against the current Imagild API contract and the public model catalog. Code, budgets and acceptance checks are illustrative; this article does not claim a new paid generation test, an uptime measurement, a search-volume study or a ranking of AI citations.
Prepared by Imagild Editorial, the company editorial team responsible for these guides. AI assisted research and drafting. This is documentation-based guidance with proposed examples, not a claim of hands-on testing or an independent product review. Report a correction.
Cover: Original Imagild product screenshot captured on 28 October 2026; model availability and prices may change. This image is not an API response or a performance benchmark.