Fetch.ai ships paid horoscope agent

Fetch.ai has published a tutorial that shows developers how to build a paid horoscope agent using Stripe, its Python-based agent tooling and the ASI:One API. The company described the example in a May 23 social-media post as a four-step build for a Stripe-integrated horoscope agent. Its documentation says the example is designed around agent transactions, payment collection and delivery of a paid response after checkout. The published example asks a user for their zodiac sign, requests payment and then returns a “horoscope of the day” after Stripe confirms the transaction, according to Fetch.ai’s Innovation Lab documentation. The guide uses ASI:One for the generated response and Fetch.ai’s agent framework for the transaction flow. (innovationlab.fetch.ai) ### What exactly did Fetch.ai ship? Fetch.ai’s Innovation Lab page lists the project as “Stripe Horoscope Agent” and says it demonstrates a seller agent that chats with users, requests a $1 payment in U.S. dollars through Stripe, and generates a horoscope after payment. The page says the payment method is Stripe and identifies the model as “asi1.” (innovationlab.fetch.ai) The company’s social post framed the release as a four-step tutorial, while the documentation points developers to a full runnable example in Fetch.ai’s `innovation-lab-examples` repository. The docs say the example is meant to keep larger code blocks out of the main documentation. ### How does the payment flow work? (innovationlab.fetch.ai) The documentation says the agent uses the Agent Payment Protocol to request payment and Stripe Checkout to process it. A separate Fetch.ai integration guide says Stripe-based agents can create payment links and use ASI:One for natural-language handling, with deployment through uAgents. (innovationlab.fetch.ai) Fetch.ai’s example breaks the flow into discrete files. The docs identify `handlers.py` as the state machine for “sign → payment → horoscope,” `stripe_payments.py` as the component that creates embedded Stripe Checkout sessions and verifies payment status, and `llm.py` as the file handling ASI:One prompts. (innovationlab.fetch.ai) ### What does a developer need to run it? Fetch.ai says developers need Python 3.11 or later, an ASI:One API key and Stripe API keys. The tutorial recommends Stripe sandbox keys first and says no real money moves in that setup. The run instructions published by Fetch.ai tell users to clone the examples repository, create a virtual environment, install requirements, copy the environment file and run `agent.py`. (innovationlab.fetch.ai) The docs describe `agent.py` as the entry point that loads environment variables and includes the chat and payment protocols. ### Where does ASI:One fit into the build? Fetch.ai’s ASI:One documentation describes the platform as an API for agentic AI, with chat completions, function calling and agent-oriented capabilities. The quickstart says developers authenticate with a bearer token and call the `api.asi1.ai` chat completions endpoint using the `asi1` model. (innovationlab.fetch.ai) The horoscope example uses that model after payment clears. Fetch.ai says `llm.py` handles “normal reply vs horoscope generation,” which places the language-model step behind the payment gate in the tutorial’s workflow. ### Where can builders find the next step? Fetch.ai’s documentation says the full runnable code lives in its `innovation-lab-examples` repository under `stripe-horoscope-agent`, and the company links the example from its Innovation Lab resources. (innovationlab.fetch.ai) The broader developer docs also point builders to uAgents, Agentverse and ASI:One quickstarts for deployment and discovery. (innovationlab.fetch.ai)