PromptShuttle
run_inference
Runs a single real LLM inference directly against a model (no flow), and returns the model's response plus token usage and cost. This EXECUTES a billed provider call and consumes tenant credits. Useful for testing a model/prompt, comparing reasoning-effort levels, or reproducing behavior. The returned runId can be passed to get_run for the full per-iteration detail. For server-side tool execution and multi-step agents, run a flow instead.
Remote Bring a key revuo:promptshuttle
Remote (network-hosted) · Needs a self-provisionable API key
Other tools also called run_inference?
See providers with this name
Input Schema
{
"type": "object",
"properties": {
"prompt": {
"description": "The user prompt / message text to send to the model.",
"type": "string"
},
"model": {
"description": "Model name, e.g. 'gpt-5.6-terra', 'claude-sonnet-4-6', 'grok-4.5'. Omit for the default (gemini-2.5-flash).",
"type": "string",
"default": ""
},
"systemPrompt": {
"description": "Optional system prompt, prepended as a system message.",
"type": "string",
"default": ""
},
"reasoningEffort": {
"description": "Reasoning effort: 'minimal', 'low', 'medium', or 'high'. Omit for the provider default. Native on OpenAI/xAI; a thinking budget on Anthropic/Gemini.",
"type": "string",
"default": ""
},
"temperature": {
"description": "Sampling temperature (0-2). Omit for the provider default. Ignored by models that don't support it.",
"type": [
"number",
"null"
],
"default": null
},
"maxTokens": {
"description": "Maximum output tokens. Omit for the model default.",
"type": [
"integer",
"null"
],
"default": null
},
"maxThinkingTokens": {
"description": "Explicit thinking-token budget (Anthropic/Gemini). Overrides the reasoningEffort-derived budget when both are set.",
"type": [
"integer",
"null"
],
"default": null
},
"vendorToolsJson": {
"description": "Optional JSON array of provider-native tool names to enable, e.g. '[\"web_search\"]'.",
"type": "string",
"default": ""
},
"responseSchemaJson": {
"description": "Optional JSON Schema object for structured output. Must be a JSON object, not a JSON-encoded string.",
"type": "string",
"default": ""
},
"environment": {
"description": "Environment label for the invocation log (default 'mcp').",
"type": "string",
"default": "mcp"
},
"isDebug": {
"description": "Capture verbatim provider request/response bodies for debugging (visible via get_run).",
"type": "boolean",
"default": false
}
},
"required": [
"prompt"
]
}