Quando colleghi questo connettore a un agente, l’agente non impara gli strumenti a memoria: se li fa dire, e li usa quando servono. Per questo un connettore aggiornato porta strumenti nuovi senza che l’agente vada rifatto.
Durante una conversazione il giro e’ sempre lo stesso: l’agente capisce dalla tua richiesta quale strumento serve e con quali valori usarlo, il gateway lo esegue e restituisce il risultato, e l’agente te lo racconta a parole. Chi compie l’operazione resti tu: l’accesso avviene con credenziali verificate, cosi’ il connettore puo’ distinguere cosa e’ permesso a te da cosa e’ permesso a un altro.
When you connect this connector to an agent, the agent doesn’t learn its tools by heart: it is told what they are, and uses them when they are needed. That is why an updated connector brings new tools along without the agent having to be rebuilt.
During a conversation the loop is always the same: the agent works out from your request which tool is needed and with which values, the gateway runs it and returns the result, and the agent tells you in plain words. You remain the one performing the operation: access happens with verified credentials, so the connector can tell what you are allowed to do from what someone else is allowed to do.
generateImageUrlgenerateImageeditImage.
{
"tool": "generateImageUrl",
"parameters": {
"prompt": "<string>"
}
}
9 strumenti. Non li chiami tu: li sceglie l’agente in base a quello che gli chiedi. Servono a farti capire cosa puoi aspettarti e quali informazioni ti verranno chieste. L’elenco effettivo puo’ variare con la configurazione del collegamento. 9 tools. You don’t call them yourself: the agent picks them based on what you ask. They are here so you can see what to expect and which details you will be asked for. The actual list may vary depending on how the connection is configured.
generateImageUrl
Generate an image URL from a text prompt. User-configured settings in MCP config will be used as defaults unless specifically overridden.
Gli serve:It needs: promptstring
Puo’ usare:It may use: model seed width height enhance safe
{
"tool": "generateImageUrl",
"parameters": {
"prompt": "<string>"
}
}
generateImage
Generate an image, return the base64-encoded data, and save to a file by default. User-configured settings in MCP config will be used as defaults unless specifically overridden.
Gli serve:It needs: promptstring
Puo’ usare:It may use: model seed width height enhance safe outputPath fileName format
{
"tool": "generateImage",
"parameters": {
"prompt": "<string>"
}
}
editImage
Edit or modify an existing image based on a text prompt. User-configured settings in MCP config will be used as defaults unless specifically overridden.
Gli serve:It needs: promptstring imageUrlstring
Puo’ usare:It may use: model seed width height enhance safe outputPath fileName format
{
"tool": "editImage",
"parameters": {
"prompt": "<string>",
"imageUrl": "<string>"
}
}
generateImageFromReference
Generate a new image using an existing image as reference. User-configured settings in MCP config will be used as defaults unless specifically overridden.
Gli serve:It needs: promptstring imageUrlstring
Puo’ usare:It may use: model seed width height enhance safe outputPath fileName format
{
"tool": "generateImageFromReference",
"parameters": {
"prompt": "<string>",
"imageUrl": "<string>"
}
}
listImageModels
List available image models
{
"tool": "listImageModels",
"parameters": {}
}
respondAudio
Generate an audio response to a text prompt and play it through the system
Gli serve:It needs: promptstring
Puo’ usare:It may use: voice seed
{
"tool": "respondAudio",
"parameters": {
"prompt": "<string>"
}
}
listAudioVoices
List all available audio voices for text-to-speech generation
{
"tool": "listAudioVoices",
"parameters": {}
}
respondText
Respond with text to a prompt using the Pollinations Text API. User-configured settings in MCP config will be used as defaults unless specifically overridden.
Gli serve:It needs: promptstring
Puo’ usare:It may use: model seed temperature top_p system
{
"tool": "respondText",
"parameters": {
"prompt": "<string>"
}
}
listTextModels
List available text models
{
"tool": "listTextModels",
"parameters": {}
}
Elenco rilevato il 11/08/2026. List observed on Aug 11, 2026.