Insightful Status Summaries using Google Gemini 1.5 Flash

Leon Nicholls
9 min readJun 13, 2024


Platforms like the Teia NFT marketplace rely on a complex symphony of APIs, blockchain nodes, and decentralized storage; even a minor glitch can have significant consequences. Downtime means lost sales, frustrated users, and a potential hit to your project’s reputation. But what if you had an LLM-powered crystal ball that could warn you of possible issues before they wreak havoc?

Enter Google Gemini, your new secret weapon in the battle for Web3 reliability. This cutting-edge language model can transform Teia’s status logs into easy-to-understand summaries, highlighting critical issues and potential trouble spots. Think of it as your personal LLM assistant, tirelessly monitoring your platform’s health and whispering warnings in your ear before disaster strikes.

This guide will explore how Google Gemini can simplify your web status monitoring. We’ll show you how to craft the perfect prompts to extract actionable insights from Teia’s logs, empowering you to keep your marketplace running smoothly even when the blockchain throws a tantrum. So buckle up and get ready to unlock the power of Google Gemini — your web project’s new best friend.

Note: This article is more techy than my previous articles about Google Gemini Advanced. However, you don’t need to be a developer to use any of the tools mentioned.

Demystifying the Teia Tech

Behind Teia’s sleek interface lies a labyrinthine network of tech wizardry. Teia is a decentralized application that doesn’t rely on centralized servers. It interacts with the Tezos blockchain and IPFS (a decentralized file storage system) to display and manage NFTs. This architecture ensures that Teia is resilient, transparent, and resistant to censorship.

The Teia UI, the part of the platform users interact with, is a React application. The codebase handles everything from browsing and searching for NFTs to minting, buying, selling, and managing collections. It’s a complex system with many moving parts, all working together to provide a seamless user experience.

And why should you care about all this techy stuff? Every piece of this puzzle must work harmoniously for Teia to function smoothly. If one part goes haywire — the connection to the Tezos blockchain hiccups or the IPFS gateway has a bad day — the whole platform could stumble. That’s why keeping a close eye on the status of all these dependencies is crucial. It’s like making sure all the cogs in a clock are turning at the right speed — otherwise, time gets a little wonky, and nobody wants that.

Note: You can read about the Teia technical design in the Teia Developer Documentation.

Enter the Answer Bot: Teia’s Status Guardian

You might wonder, “How does Teia keep tabs on this intricate web of dependencies?” Please enter the Answer Bot, Teia’s trusty sidekick and resident status guru, which I introduced in a previous post. This little bot lives on Teia’s Discord server and monitors Teia’s vital signs.

Think of the Answer Bot as Teia’s health tracker. Every minute, it diligently checks on all the crucial components that make Teia tick. It pings the Tezos blockchain, ensures the indexer is humming, and checks if the IPFS gateway is playing nicely. It’s like a doctor doing a routine checkup, but instead of listening to your heartbeat, it’s checking the pulse of a decentralized marketplace.

But the Answer Bot doesn’t just keep this information to itself. It meticulously records its findings in a log, like a digital diary of Teia’s health. Each entry in this log is a snapshot of a specific moment in time, detailing the status of each dependency. It’s a bit like those old-school ticker tapes that used to display stock prices, but instead of numbers, you’ve got a series of “OKs, “WARNING”s, or “ERROR”s.

You might think, “Great, another log file to decipher.” But fear not! This is where Google Gemini comes in to work its magic. You see, Gemini thrives on structured data, and the Answer Bot’s log is just that — a treasure trove of information waiting to be unlocked. By feeding these logs into Gemini, we can transform them from a jumble of technical jargon into a clear, concise summary of Teia’s health. It’s like having a translator who speaks both “bot” and “human,” making communication between you and your Web3 project much more effortless.

Note: The code for the status tracking is open-sourced.

Summoning Your Gemini Wizard: Crafting the Perfect Prompt

We’ve got our trusty Answer Bot diligently logging Teia’s status. Now, let’s unleash the true power of Google Gemini. Think of this as casting a spell, but instead of arcane incantations, we’re using the art of prompt engineering. This is where we tell Gemini precisely what we want it to do with those logs.

We’ll tap into the power of Google Gemini 1.5 Flash, the speedy Gonzales of language models. Think of it as the Usain Bolt of Google’s Gemini family — built for speed and efficiency without breaking the bank. It’s a champ at handling tons of data simultaneously and has this mind-blowing “one-million-token context window.” In plain English, it can juggle a massive amount of information, making it perfect for tasks like chatbots, summarizing long documents, writing captions, and even extracting data like a pro.

Step 1: Roleplay Time!

First, we need to set the stage. We’ll instruct Gemini to “Act as a Site Reliability Engineer for a Web3 marketplace.” This is for more than just fun (though it is a bit fun). By assigning Gemini this role, we’re tapping into its knowledge of technical systems and troubleshooting. It’s like giving it a crash course in Web3 infrastructure, so it knows what to look for in the logs.

Step 2: Spill the Tea (aka Context)

Next, we need to give Gemini the lowdown on Teia. We’ll explain that it’s an NFT marketplace built on Tezos and list all the crucial dependencies it relies on. This is like giving Gemini a map of Teia’s inner workings so it knows where to focus its attention. We also explain which technical dependencies are essential and which are less so.

Step 3: Make a Wish List

Now for the fun part — telling Gemini precisely what we want. Here’s a sample wish list you can use:

  • “Summarize the status of the dependencies concisely and clearly, and identify key trends and takeaways.” This tells Gemini to give us the big picture, focusing on the most essential stuff.
  • “Include all relevant details in your summary and analysis.” This helps us spot potential problems before they become major headaches.
  • “Keep the response non-technical.” We want our summaries accessible to everyone on the team, not just the tech wizards.
  • “Title the summary as: ‘Teia Daily Status Summary.’” This helps us keep things organized.

Step 4: Unleash the Data

Finally, we’ll give Gemini the Answer Bot’s logs. This is like giving it the raw ingredients for a potion — it’s up to Gemini to mix them and create something magical. All the summaries for the past day are combined into a list, with the date of each summary used as a separator.

With these four simple steps, you’ve summoned your Gemini wizard and set it on a quest to transform your status logs into valuable insights. It’s like having a personal oracle that can predict the future of your Web3 project, helping you stay one step ahead of any potential problems.

Note: You can see the complete prompt passed onto the Gemini API.

Gemini in Action: Teia’s Daily Digest

Drumroll, please! It’s time to unveil the fruits of Gemini’s labor: the Teia Daily Status Summary. Imagine this as your personalized newspaper, hot off the digital press, delivering the most critical news about your Web3 marketplace.

A Glimpse into Gemini’s Brilliance

Let’s say the Answer Bot has diligently collected data on Teia’s dependencies for a whole day. There have been a few hiccups here and there. The TzKT indexer may be sluggish, or the IPFS gateway had a brief timeout. Instead of sifting through pages of logs yourself, you ask Gemini to work its magic.

Abracadabra! Gemini conjures up a summary that looks something like this:

Making Sense of the Magic

See how Gemini distills a day’s technical mumbo-jumbo into a compact report? It’s like having a personal news anchor who only reports the most relevant headlines. You instantly know what’s working, what’s not, and what needs your attention.

From Insights to Action

Armed with this summary, the Teia team can now spring into action. They can investigate the TzKT indexer delays by optimizing queries or adding more resources. They can also closely monitor the IPFS gateway, ensuring it stays responsive and exploring backup options.

The beauty of Gemini’s summary is that it empowers everyone on the team, regardless of their technical background. Even non-technical folks can understand the key takeaways and contribute to the discussion. It’s like having a universal translator for your Web3 project, fostering collaboration and ensuring everyone is on the same page.

Leveling Up: Advanced Gemini Prompt Engineering — Become a Prompt Whisperer

Think you’ve mastered the art of Gemini summoning? Buckle up because we’re about to dive into the deep end of prompt engineering. This is where things get interesting, and you can truly unleash the full potential of your LLM oracle.

Iteration: The Path to Prompt Perfection

Remember, Rome wasn’t built in a day, nor is the perfect Gemini prompt. It’s all about iteration, baby! Start with a basic prompt, see what kind of summaries Gemini conjures up, and then tweak it based on the results. You may need to be more specific in your requests, or you may want to focus on different aspects of Teia’s status. It’s like tuning a musical instrument — keep adjusting until you hit the sweet spot.

Customization: Tailor-made for Teia

One size doesn’t fit all in the world of prompt engineering. The Teia marketplace is unique, so why settle for a generic prompt? Get creative and customize your prompts to suit your specific needs. Want to zero in on critical errors? Are you curious about the performance of particular components? Tell Gemini to give you the juicy details. The possibilities are endless!

Experimentation: The Mad Scientist Approach

Feel free to get your hands dirty and experiment with different prompt variations. Try adding new instructions, changing the wording, or even throwing in some emojis for fun (who knows, Gemini might appreciate the flair!). The more you experiment, the better you’ll understand how to communicate with your LLM oracle and get the insights you crave.

Beyond Summaries: Gemini’s Hidden Talents

While daily status summaries are handy, Gemini has a few more tricks. With the proper prompts, you can unlock its potential for:

  • Root Cause Analysis: Gemini can help you examine the logs more closely to uncover the underlying causes of errors or performance issues.
  • Outage Prediction: By analyzing historical data, Gemini can predict potential outages before they happen, giving you time to prepare.
  • Performance Optimization: Gemini can identify bottlenecks and suggest ways to improve Teia’s overall performance.

The key is to think outside the box and explore Gemini’s vast possibilities. You may discover a hidden talent that revolutionizes managing your Web3 project.

Note: If you are not a developer, you can prototype a prompt using the free Google AI Studio instead of the Gemini API.


So, there you have it, folks! We’ve journeyed through the mystical land of Teia, explored its intricate tech stack, and summoned the mighty Google Gemini to wrangle its status logs into submission. We’ve learned how to craft the perfect prompt, decipher Gemini’s technical summaries, and even dabbled in advanced prompt engineering.

Remember, Gemini is more than just a status summarizer. It’s a powerful tool that can help you gain deeper insights into your platform, predict potential issues, and optimize performance. So go forth, brave adventurer, and explore the endless possibilities in Google Gemini.

Check out my reading list of other Google Gemini articles.

This post was created with the help of AI writing tools, carefully reviewed, and polished by the human author.