Skip to content

Power Puter: add Civitai filter for Lora triggers#679

Open
kaalibro wants to merge 1 commit into
rgthree:mainfrom
comfy-nodes:feat/power-puter-triggers-civitai
Open

Power Puter: add Civitai filter for Lora triggers#679
kaalibro wants to merge 1 commit into
rgthree:mainfrom
comfy-nodes:feat/power-puter-triggers-civitai

Conversation

@kaalibro

@kaalibro kaalibro commented Feb 6, 2026

Copy link
Copy Markdown

Added optional civitai parameter to get_enabled_triggers_from_prompt_node() method in Power Lora Loader to filter trigger words by their source.

Changes

File: py/power_lora_loader.py

  • Added optional civitai: bool = False parameter to get_enabled_triggers_from_prompt_node()
  • When civitai=True, the method now filters trigger words to only include those with "civitai": true in the model info JSON
  • When civitai=False (default), all trigger words are returned as before

Usage in Power Puter node

# Get all trigger words (default behavior)
triggers(max_each = None)

# Get only CivitAI-sourced trigger words
triggers(max_each = None, civitai = True)

Example Model Info JSON

{
  "trainedWords": [
    {
      "word": "some_trigger_word",
      "count": 42,
      "metadata": true
    },
    {
      "word": "high_detail",
      "count": 42,
      "metadata": true,
      "civitai": true
    }
  ]
}

With civitai=True, only "high_detail" would be returned; with default behavior, both words would be returned.

This change allows users to selectively include only trigger words that were fetched from CivitAI, which can be useful for filtering out manually added or less reliable trigger words.

The `get_enabled_triggers_from_prompt_node` function now accepts a `civitai`
parameter. When set to `True`, only trigger words explicitly marked as
`civitai` in the Lora info will be returned.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant