AI Actions

General

prompt

Sends a prompt to an AI provider and returns the generated response text.

Supported providers: deepseek, openai, anthropic. frequencyPenalty and presencePenalty are only supported by OpenAI and DeepSeek.

Returns: The AI-generated response text.

Parameter	Type	Required	Default	Description
provider	deepseek | openai | anthropic	✓	—	AI provider to use (deepseek, openai, or anthropic).
apiKey	string	✓	—	API key for the selected provider.
prompt	string	✓	—	The user prompt to send to the AI model.
model	string		—	Model identifier (e.g. gpt-4o, claude-3-5-sonnet). Defaults to provider default.
systemMessage	string		—	Optional system/instruction message to set model behavior.
temperature	number		—	Sampling temperature (0–2). Higher values produce more random output.
maxTokens	number		—	Maximum number of tokens to generate (1–100000).
topP	number		—	Nucleus sampling cutoff (0–1). Alternative to temperature.
stream	boolean		—	Whether to stream the response token by token.
frequencyPenalty	number		—	Penalty for token frequency (-2 to 2). Reduces repetition.
presencePenalty	number		—	Penalty for token presence (-2 to 2). Encourages new topics.

# Summarize a document with DeepSeek using a system message and moderate temperature
- group: ai
  name: prompt
  params:
    provider: deepseek
    apiKey: '{{ deepseekKey }}'
    prompt: 'Summarize the following document: {{ documentText }}'
    systemMessage: 'You are a concise technical writer. Return plain text only.'
    temperature: 0.5
    maxTokens: 300