How to upload and index files

Use this guide when a project needs document context before analysis.

1. Open the target project or workspace

Choose the project that should own the files. File context is project-scoped, so upload into the same project where you plan to run the workflow.

2. Start the upload from the file context area

Use the project or workspace file panel to add files. Under the hood, the product uses a presign-upload-confirm flow, but the user-facing task is simply to attach the files from the application.

3. Wait for indexing to finish

After upload confirmation, the files move through indexing states. The key states exposed in product and API docs are:

• PENDING
• INDEXING
• READY
• FAILED

4. Use only READY files as context

Do not treat PENDING, INDEXING, or FAILED files as usable run context. The runtime path is strict about this and only valid READY files should be selected for supported runs.

5. Retry or replace failed files

If a file reaches FAILED, treat it as unusable context. Re-upload or investigate the file before depending on it in a workflow.

6. Move to analysis

Once the needed files are READY, continue with How to run SR&ED analysis.

For the exact upload contract, see File upload and indexing.