Save To Zip¶
Collects and organizes multiple node outputs into a structured folder and creates a ZIP archive for download. Inputs can be connected progressively (results_1, results_2, ...) and are grouped into folders based on metadata or source node, with optional seeds/configs included.
Usage¶
Use this node at the end of a workflow to bundle outputs from various nodes into a single downloadable ZIP. Connect any output to results_1, then chain additional outputs to results_2, results_3, etc. The node will automatically structure files into folders, include seeds/configs if provided by upstream nodes, and generate a unique archive per workflow execution.
Inputs¶
| Field | Required | Type | Description | Example |
|---|---|---|---|---|
| results_1 | True | WILDCARD | First output to include in the ZIP. Accepts any output from any node. Enables the next input when connected. | Any data type, e.g., a string, number, dict of files like {'model.pdb': '...'}, or other node-specific outputs. |
| results_2 | False | WILDCARD | Second output to include in the ZIP. Appears only after results_1 is connected. | Any additional output to save. |
Outputs¶
| Field | Type | Description | Example |
|---|---|---|---|
| url | STRING | Path or URL to the output location containing the generated ZIP archive for this workflow execution. | /path/to/output/ |
Important Notes¶
- Connect outputs in order: results_1 must be connected before results_2 becomes available, and so on up to results_101.
- Inputs that carry Salt metadata (folder_name, file_name, seed, config) are saved using that structure; seeds and configs (if present) are aggregated into seeds.json and configs.json at the root.
- If an input contains multiple sub-files (e.g., PDB or A3M collections), the node writes them as separate files under a folder matching the main file name.
- Inputs without metadata are placed into a folder named after the source node class, with file names like output_
.txt. - The generated archive is uniquely named using the workflow_execution_id (and the index of this Save node instance) to avoid collisions.
- The output location is environment-dependent; the runtime may expose a downloadable link in the Output node.
Troubleshooting¶
- No files in ZIP or missing expected files: Ensure each results_i is connected and producing a value at execution time.
- Unexpected folder or file naming: Check if upstream nodes provide Salt metadata (folder_name, file_name). Without metadata, generic names are used.
- Download link not clickable or missing: Connect the 'url' output to an Output node or the designated display node for links.
- Permission or write errors: Verify the runtime has write access to the configured output directory.
- Invalid or complex data structures not saved as JSON: Non-serializable data is saved as plain text; consider converting outputs to strings or simple JSON-serializable objects upstream.
Example Pipelines¶
