One of the main challenges I faced during development was the limited flexibility and customization options offered by Ohara.ai. While it’s a powerful tool with seamless Base network integration and support for certain Web3 tools like Thirdweb Contracts and the 0x API, I often felt constrained when trying to go beyond its built-in capabilities.

For instance, I needed to fetch data from third-party APIs like GeckoTerminal (e.g., https://api.geckoterminal.com/api/v2/networks/base/trending_pools?page=1), but Ohara.ai didn’t support direct API calls to such external services. To overcome this, I had to explore creative workarounds—such as pre-fetching and storing data in resource files or manipulating the input/output through prompt engineering to simulate real-time data interactions.

Similarly, when working with influencer data, I had to import static datasets manually and optimize my prompts extensively to achieve the desired outputs. These constraints pushed me to experiment and innovate within the platform’s limits, enhancing my adaptability and problem-solving skills.

Despite these challenges, I ohara.ai quite impressive, and with some enhancements to API extensibility, it could evolve into a more robust and developer-friendly platform.