Web3gpt

Web3GPT Skill

Use the Web3GPT skill endpoint to start a chat, continue a chat, and optionally return the full history.

Base endpoint:

https://w3gpt.ai/api/skill

Start a new chat

Call the endpoint with no parameters.

curl https://w3gpt.ai/api/skill

Response:

{
  "agentId": "agent_web3gpt",
  "chatId": "generated-chat-id",
  "response": null
}

Save the chatId. That is the secret for continuing the same thread.

Continue a chat

Send a message with the same chatId.

curl -X POST "https://w3gpt.ai/api/skill?chatId=generated-chat-id" \
  -H "Content-Type: application/json" \
  -d '{"message":"Deploy an ERC20 on Polygon mainnet with 1,000,000 supply"}'

Response:

{
  "agentId": "agent_web3gpt",
  "chatId": "generated-chat-id",
  "response": "..."
}

Full history

History is off by default.

• history=true returns the full simplified history
• history=false or omitting it returns only the latest response
• full=true is supported as an alias for history=true

Example:

curl "https://w3gpt.ai/api/skill?chatId=generated-chat-id&history=true"

Optional parameters

• chatId: continue an existing chat
• agentId: use a different Web3GPT agent
• message: send a message by query string for small requests
• history: include the full history, defaults to false
• full: alias for history

Notes

• If you call the endpoint without a chatId, a new chat is created automatically.
• If you call the endpoint without a message, you still get back a chatId.
• Deployments happen through the agent conversation. Tell the agent which chain to use.
• Polygon mainnet deployment is available through the agent/skill endpoint, not the wallet UI connectors.

Source: Markeljan/web3gpt — distributed by TomeVault.