VLM System Query Add Presets:
This node helps you assemble text for a vision-language model (VLM) query using a preset. It places your wording between preset-defined text that comes before and after it, giving you a convenient way to reuse a consistent query structure while changing the central request. The node only combines strings; it does not send a request to a model or check how a model will interpret the result.
VLM System Query Add Presets Input Parameters:
preset
Selects which preset-defined text to place around your wording. The available options are generated from the names in the additional VLM system-query preset collection, so the exact choices depend on the presets installed with your node package. Each option can supply text before your wording, after it, or both; if either part is absent, that part is treated as empty. The default is the first available option in alphabetical order, or an empty value if there are no options. Minimum and maximum values do not apply to this selection.
text
The wording you want to insert between the selected preset’s beginning and ending text. This is a multiline string field, so you can write a short request or a longer passage. Its default is an empty string, and the node applies no length limit in its source code. Leaving it blank produces only the preset-defined text, if any; changing it changes the central portion of the assembled result.
VLM System Query Add Presets Output Parameters:
system_instruction
A string containing the selected preset’s beginning text, followed immediately by your text, followed immediately by the preset’s ending text, in that order. This is the assembled text you can pass to another node that accepts a string. The node adds no separators of its own, so any spaces or line breaks between the parts must come from your wording or the preset content. Despite the output’s name, this node does not verify that the receiving model or node will treat the string as a system instruction.
VLM System Query Add Presets Usage Tips:
- Choose a preset whose assembled wording suits your task, then put the part you want to change between runs in
text. This lets you reuse the same surrounding text without retyping it. - If the result appears to run words together, add a space or line break at the start or end of
textas needed. The node joins the three parts directly. - For a custom multiline request, enter the full passage in
text; the node preserves your supplied wording when it assembles the string.
VLM System Query Add Presets Common Errors and Solutions:
No preset options appear
- Explanation: The preset menu is built from entries in the additional VLM system-query preset collection. If that collection has no matching entries, there are no names to show.
- Solution: Check that the node package and its preset data are installed and loading correctly. Restore or update the package if its preset data is missing.
Invalid preset selection
- Explanation: A saved workflow may refer to a preset that is no longer among the available options, for example after the node package’s presets change. The node itself does not define a specific error message for this situation; workflow validation may report the invalid selection.
- Solution: Open the
presetmenu and choose one of the currently available options, then save the workflow again.
Output is blank or missing expected preset text
- Explanation: The node uses an empty string when a selected preset’s beginning or ending entry is missing. If
textis also empty, the entire result can be blank. - Solution: Enter wording in
textand check the selected preset. If you expected surrounding text, verify that the preset data includes the corresponding beginning or ending entry.
Words or lines run together
- Explanation: The node joins preset text and
textwithout inserting spaces or line breaks. - Solution: Add the needed spacing or line breaks to
text, or choose a preset whose supplied text already provides the separation you need.
