Replit Agent 3 | (Tested)

Prompt Engineering Guide · Intermediate ·🤖 AI Agents & Automation ·9mo ago

Key Takeaways

Replit Agent 3 is tested and explored in this video, focusing on AI agents and their applications in coding and programming. The video is part of the Prompt Engineering Guide series, which offers an academy for learning about AI agents.

Full Transcript

Replet recently released agent tree. So you can see the announcement here once you go to replet.com. I've been using Replet to build all sorts of small little applications using their replet agent feature. With this agent tree comes this release agents and automations which enables you to build AI bots or workflow automations in plain English. So basically you vibe code your AI automations and they're all powered by agents. This is a better feature, but it has options such as building apps around Slack or Telegram integration. It even has a timebased one, which is something that I built and I want to show you a little example of that. And then there is this event based one that's coming soon. What you see here is the playground and this is an automation workflow that I already built and it's powered by an agentic framework which in turn is powered by open router models. So I'm using the open router API and this is a timebased trigger which means this automation is going to run every four hours once I publish it. So I haven't really publish it because I'm just testing it. And so the idea here is that you can use rapid agent. You know you vibe code the automation. It sets up the database if it needs it. It configures the agent framework that you need to power the automation. It also allows you to incorporate your API keys if you're using specific models which in this case I'm using open router. So all of that stuff happens within replet agent. So replet agent is not open for me here because I just wanted to show you the end result of this workflow that I quickly build. So this automation workflow I can already test it and then it has like a couple of steps. It's a little side project that I've been wanting to build for a few weeks now. And basically what I want is I want to be able to check the change log of cloud code and receive updates every 4 hours if something change in the code. I could do it every 24 hours. I could do whatever schedule I want because that's the point of the automation workflow. And the way how I have it set up here it will fetch the change log and you can see that this is powered by code. It's all code and I didn't generate one line of code here. Everything was done by replet agent. And then here is just formatting change log. So it uses AI agent to format and summarize change log updates for Slack. So this is something I want to experiment with. For now it's quite basic. It's just pulling information then summarizing it and sending it to Slack. So it sends a notification and then it updates the state. So it's using a little database to track the change log essentially. If I check every 4 hours or if I check every 12 hours and there's no changes, then this automation is just going to send me a very simple message such as no updates. But if it does see changes, then that will show up. I have so many ideas for this, but this is just V1. So now, let me show you how it works. So I'm going to test the automation here. So now it's running, and you can also view the detailed traces if you want. I really like this because automations, they usually can go wrong in so many ways. So it's good to have that observability functionality there. All right. So it looks like all of the steps completed, right? We fetch the change log. We format the change log with an AML. We send the notification to Slack. And then there's this update here that's tracked by my little database. Let's look at where this was sent. I'm going to open up Slack and show you the channel that I've configured to receive these change log notifications. So you can see here that I received the change log update. So for cloud code there is the latest update which is v 1.0.113. These were the latest updates. Deprecated pipe input in interactive mode transcript toggle key binding move to control. So and then it gives me a link here that I can go and then check the change log. The change log in question is this one from cloud code. This is where all the updates happen. For now I have to come here manually. So the problem that I'm trying to solve is to gather all of these little updates and do it on a schedule, right? every 12 hours, every 24 hours, so I never miss any important updates inside of cloud code. And this is just the first step. I have so many ideas for this. I'm actually thinking of extending the automation to be able to do more like deep research and other things like that. And because I have built this within replet, I can just expand the code or keep coding and adding more functionalities to it. I can customize it however I want. I like the power of doing that and I think building automations like that in Replet has huge potential for building useful applications like this. Let me know what you think. I'll keep experimenting with this and share more if I find any interesting insights. I'll also be doing a more extended overview with all the steps of how to set something like this up inside of Reply with all the integrations into Slack and so on. I'll be doing that for my Academy Pro members. So stay tuned for that. Thank you for watching and I'll see you all in the next

Original Description

Learn how to build with AI Agents in my academy: https://dair-ai.thinkific.com/ Use code YOUTUBE20 to get an extra 20% off. -- Tested Replit Agent 3. #ai #coding #programming
Watch on YouTube ↗ (saves to browser)
Sign in to unlock AI tutor explanation · ⚡30

Playlist

Uploads from Elvis Saravia · Elvis Saravia · 0 of 60

← Previous Next →
1 101 ways to solve search (by Pratik Bhavsar)
101 ways to solve search (by Pratik Bhavsar)
Elvis Saravia
2 TLDR Generation of Scientific Documents | ML Interview #1 with Isabel Cachola
TLDR Generation of Scientific Documents | ML Interview #1 with Isabel Cachola
Elvis Saravia
3 Sentiment Analysis: Key Milestones, Challenges and New Directions
Sentiment Analysis: Key Milestones, Challenges and New Directions
Elvis Saravia
4 Discriminative Adversarial Search for Abstractive Summarization (by Thomas Scialom)
Discriminative Adversarial Search for Abstractive Summarization (by Thomas Scialom)
Elvis Saravia
5 Question Understanding: COVID-Q: 1,600+ Questions about COVID-19
Question Understanding: COVID-Q: 1,600+ Questions about COVID-19
Elvis Saravia
6 Getting Started with NLP
Getting Started with NLP
Elvis Saravia
7 Building tools and frameworks for large-scale social media mining (by Dr. Juan M. Banda)
Building tools and frameworks for large-scale social media mining (by Dr. Juan M. Banda)
Elvis Saravia
8 TextAttack: A Framework for Data Augmentation and Adversarial Training in NLP
TextAttack: A Framework for Data Augmentation and Adversarial Training in NLP
Elvis Saravia
9 Dive into Deep Learning (Study Group): Introduction to Deep Learning | Session 1
Dive into Deep Learning (Study Group): Introduction to Deep Learning | Session 1
Elvis Saravia
10 Dive into Deep Learning (Study Group): Multilayer Perceptrons | Session 4
Dive into Deep Learning (Study Group): Multilayer Perceptrons | Session 4
Elvis Saravia
11 How I read and annotate ML papers
How I read and annotate ML papers
Elvis Saravia
12 Keep Learning ML  (Session 1) | DSV, CompLex, Modern tools for emotions
Keep Learning ML (Session 1) | DSV, CompLex, Modern tools for emotions
Elvis Saravia
13 Dive into Deep Learning (Study Group): Preliminaries | Session 2
Dive into Deep Learning (Study Group): Preliminaries | Session 2
Elvis Saravia
14 Keep Learning ML #2 | Language-conditioned policy learning, Effective ML Testing, EagerPy
Keep Learning ML #2 | Language-conditioned policy learning, Effective ML Testing, EagerPy
Elvis Saravia
15 Dive into Deep Learning (Study Group): Linear Neural Networks | Session 3
Dive into Deep Learning (Study Group): Linear Neural Networks | Session 3
Elvis Saravia
16 Dive into Deep Learning (Study Group): Multilayer Perceptrons | Session 4
Dive into Deep Learning (Study Group): Multilayer Perceptrons | Session 4
Elvis Saravia
17 Keep Learning ML #3 | Contrastively Trained Structured World Models
Keep Learning ML #3 | Contrastively Trained Structured World Models
Elvis Saravia
18 Dive into Deep Learning (Study Group): Deep Learning Computation with PyTorch |  Session 5
Dive into Deep Learning (Study Group): Deep Learning Computation with PyTorch | Session 5
Elvis Saravia
19 Dive into Deep Learning (Study Group): Convolutional Neural Networks | Session 6
Dive into Deep Learning (Study Group): Convolutional Neural Networks | Session 6
Elvis Saravia
20 Dive into Deep Learning (Study Group): Modern CNNs | Session 7
Dive into Deep Learning (Study Group): Modern CNNs | Session 7
Elvis Saravia
21 101 ways to solve neural search with Jina
101 ways to solve neural search with Jina
Elvis Saravia
22 (Hopefully-Reusable) Life Lessons for PhD Students in NLP
(Hopefully-Reusable) Life Lessons for PhD Students in NLP
Elvis Saravia
23 How to save the world and forward your career in 5 easy steps | Women in NLP Talks
How to save the world and forward your career in 5 easy steps | Women in NLP Talks
Elvis Saravia
24 Prompt Engineering Overview
Prompt Engineering Overview
Elvis Saravia
25 Getting Started with the OpenAI Playground
Getting Started with the OpenAI Playground
Elvis Saravia
26 LM-Guided Chain of Thought
LM-Guided Chain of Thought
Elvis Saravia
27 Elements of a Prompt
Elements of a Prompt
Elvis Saravia
28 Reasoning with Intermediate Revision and Search with LLMs #chatgpt #ai #llms #science #programming
Reasoning with Intermediate Revision and Search with LLMs #chatgpt #ai #llms #science #programming
Elvis Saravia
29 General Tips for Designing Prompts
General Tips for Designing Prompts
Elvis Saravia
30 Efficient Infinite Context Transformers #ai #machinelearning #research #llms #science
Efficient Infinite Context Transformers #ai #machinelearning #research #llms #science
Elvis Saravia
31 Best Practices and Lessons Learned on Synthetic Data for Language Models #ai #machinelearning #genai
Best Practices and Lessons Learned on Synthetic Data for Language Models #ai #machinelearning #genai
Elvis Saravia
32 Reducing Hallucinations in Structured Outputs via RAG #chatgpt #ai #llms #programming
Reducing Hallucinations in Structured Outputs via RAG #chatgpt #ai #llms #programming
Elvis Saravia
33 Basic Prompt Examples for LLMs
Basic Prompt Examples for LLMs
Elvis Saravia
34 LLM In Context Recall is Prompt Dependent  #llms #ai #chatgpt #machinelearning
LLM In Context Recall is Prompt Dependent #llms #ai #chatgpt #machinelearning
Elvis Saravia
35 Zero-shot Prompting Explained
Zero-shot Prompting Explained
Elvis Saravia
36 RAG Faithfulness #llms #ai #gpt4
RAG Faithfulness #llms #ai #gpt4
Elvis Saravia
37 Understanding LLM Settings
Understanding LLM Settings
Elvis Saravia
38 Llama 3 is here! | First impressions and thoughts
Llama 3 is here! | First impressions and thoughts
Elvis Saravia
39 Llama 3 is Here! #ai #llms #llama3
Llama 3 is Here! #ai #llms #llama3
Elvis Saravia
40 Microsoft introduces Phi-3 | The most capable small language model?
Microsoft introduces Phi-3 | The most capable small language model?
Elvis Saravia
41 Microsoft introduces Phi-3! #ai #llms #microsoft
Microsoft introduces Phi-3! #ai #llms #microsoft
Elvis Saravia
42 Make Your LLM Fully Utilize the Context #ai #llms #machinelearning
Make Your LLM Fully Utilize the Context #ai #llms #machinelearning
Elvis Saravia
43 When to Retrieve? #ai #llms #machinelearning
When to Retrieve? #ai #llms #machinelearning
Elvis Saravia
44 Training an LLM to effectively use information retrieval
Training an LLM to effectively use information retrieval
Elvis Saravia
45 State-of-the-art open-source LLM judges #ai #machinelearning #gpt4
State-of-the-art open-source LLM judges #ai #machinelearning #gpt4
Elvis Saravia
46 Better and Faster LLMs via Multi-token Prediction
Better and Faster LLMs via Multi-token Prediction
Elvis Saravia
47 AlphaMath Almost Zero #ai #science #machinelearning
AlphaMath Almost Zero #ai #science #machinelearning
Elvis Saravia
48 SWE-Agent | An LLM-based Software Engineering Agent
SWE-Agent | An LLM-based Software Engineering Agent
Elvis Saravia
49 [LLM NEWS] AlphaFold 3, xLSTM, OpenAI's Model Spec, DeepSeek-V2, OpenDevin CodeAct 1.0
[LLM NEWS] AlphaFold 3, xLSTM, OpenAI's Model Spec, DeepSeek-V2, OpenDevin CodeAct 1.0
Elvis Saravia
50 LLM-powered tool for web scraping #ai #chatgpt #engineering
LLM-powered tool for web scraping #ai #chatgpt #engineering
Elvis Saravia
51 Learn about LLMs in this NEW course #ai #chatgpt #engineering
Learn about LLMs in this NEW course #ai #chatgpt #engineering
Elvis Saravia
52 [LLM NEWS] KANs, Gemma 10M Context, OpenAI Updates?, Automatic Prompt Engineering, Tokenizer Arena
[LLM NEWS] KANs, Gemma 10M Context, OpenAI Updates?, Automatic Prompt Engineering, Tokenizer Arena
Elvis Saravia
53 [LLM News] GPT4-o, Project Astra, Veo, Copilot+ PCs, Gemini 1.5 Flash, Chameleon
[LLM News] GPT4-o, Project Astra, Veo, Copilot+ PCs, Gemini 1.5 Flash, Chameleon
Elvis Saravia
54 Enhancing Answer Selection in LLMs #ai #machinelearning #engineering
Enhancing Answer Selection in LLMs #ai #machinelearning #engineering
Elvis Saravia
55 On exploring LLMs #ai #promptengineering #chatgpt
On exploring LLMs #ai #promptengineering #chatgpt
Elvis Saravia
56 Transformers Can Do Arithmetic with the Right Embeddings #ai #machinelearning #engineering
Transformers Can Do Arithmetic with the Right Embeddings #ai #machinelearning #engineering
Elvis Saravia
57 [LLM News] xAI Series B, Codestral, LLM Guide, AutoGen Course, Symbolic Chain-of-Thought
[LLM News] xAI Series B, Codestral, LLM Guide, AutoGen Course, Symbolic Chain-of-Thought
Elvis Saravia
58 PR-Agent #ai #gpt4 #software
PR-Agent #ai #gpt4 #software
Elvis Saravia
59 Extracting features from Claude 3 Sonnet
Extracting features from Claude 3 Sonnet
Elvis Saravia
60 Has prompt engineering been solved?
Has prompt engineering been solved?
Elvis Saravia

This video tests and explores Replit Agent 3, providing insights into AI agents and their applications in coding and programming. Viewers can learn how to build and use AI agents with Replit. The Prompt Engineering Guide academy offers further learning opportunities.

Key Takeaways
  1. Test Replit Agent 3
  2. Explore AI agent applications in coding and programming
  3. Apply prompt engineering techniques
  4. Use Replit for AI agent development
  5. Enroll in the Prompt Engineering Guide academy
💡 Replit Agent 3 can be used for AI agent development and prompt engineering applications.

Related AI Lessons

I Found Firecrawl Too Expensive for My AI Agent's Knowledge Base, So I Built My Own
Learn how to build a custom knowledge base for AI agents when existing tools like Firecrawl are too expensive, and why this matters for cost-effective AI development
Dev.to · Samuel Raphael
How Managing 500+ Employees Led Me to Build WorkforceIQ: The AI Platform I Wish I Had
Learn how managing a large workforce led to the creation of WorkforceIQ, an AI platform for workforce management, and discover its key features and benefits
Medium · Startup
AI-First MVP Development: How Startups Should Build Products in 2026
Learn how AI-first MVP development can revolutionize startup product building in 2026
Dev.to · Nasif Sid
The Three Pillars of AI Agents: Platform, LLM, and Harness
Learn the three pillars of AI agents: platform, LLM, and harness, to understand the foundation of AI development
Medium · LLM
Up next
Building Great Agent Skills: The Missing Manual
AI Engineer
Watch →