<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>Forem: Guilherme Vilas Boas Roberto</title>
    <description>The latest articles on Forem by Guilherme Vilas Boas Roberto (@guintoki).</description>
    <link>https://forem.com/guintoki</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F2798599%2F595904a4-f1d0-45c0-914e-1fcbf49ab41c.jpg</url>
      <title>Forem: Guilherme Vilas Boas Roberto</title>
      <link>https://forem.com/guintoki</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://forem.com/feed/guintoki"/>
    <language>en</language>
    <item>
      <title>GitHub Copilot Agent Mode: An Even Smarter Code Assistant</title>
      <dc:creator>Guilherme Vilas Boas Roberto</dc:creator>
      <pubDate>Tue, 04 Mar 2025 18:42:04 +0000</pubDate>
      <link>https://forem.com/guintoki/github-copilot-agent-mode-an-even-smarter-code-assistant-4e3o</link>
      <guid>https://forem.com/guintoki/github-copilot-agent-mode-an-even-smarter-code-assistant-4e3o</guid>
      <description>&lt;p&gt;Recently, GitHub Copilot’s Agent Mode became available in the VSCode Insiders preview. This new mode can iterate over its own code, recognize errors, and correct them automatically. Additionally, it can suggest terminal commands and request their execution, as well as analyze runtime errors and apply auto-corrections.&lt;br&gt;
&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fupntcimc50ih4vlbihv7.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fupntcimc50ih4vlbihv7.png" alt="Agent Mode Example" width="382" height="389"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Now, instead of just executing the requested task, Copilot can also add additional tasks that were not specified in the prompt but are necessary for the solution to work correctly. It can identify its own errors and apply fixes directly in the code.&lt;/p&gt;

&lt;p&gt;I've been testing Agent Mode for a few weeks in experimental projects to better understand how it works and how it can impact my productivity in development.&lt;/p&gt;

&lt;h3&gt;
  
  
  A Practical Example
&lt;/h3&gt;

&lt;p&gt;One of the most interesting features of Agent Mode is its ability to analyze an entire project (or just specific files) and automatically generate solutions. It can even create new files when necessary.&lt;/p&gt;

&lt;p&gt;To test this, I created a simple CRUD project and asked Copilot to add a new route to list trucks and drivers together, combining two existing routes:&lt;br&gt;
&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2F0snuyune8noxqwiesaz1.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2F0snuyune8noxqwiesaz1.png" alt="Copilot Chat Request" width="375" height="128"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Agent Mode started analyzing the code, identified relevant references within the project, and determined the ideal approach to apply the requested solution.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fie28l146xbljzwba01nm.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fie28l146xbljzwba01nm.png" alt="Copilot's Response" width="374" height="603"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Shortly after, it created a new file, automatically including the necessary component imports for the listing and the requests.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Ff9m2rohg29t3fl3tcyp1.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Ff9m2rohg29t3fl3tcyp1.png" alt="Copilot's Response" width="374" height="226"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Next, it added the request to list trucks and drivers, including the filtering logic:&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fm2pfpwy5afbqjitt3bt2.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fm2pfpwy5afbqjitt3bt2.png" alt="New Component Code" width="800" height="699"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;And structured the display of results:&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2F58alstxyiwa2hg28s0y2.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2F58alstxyiwa2hg28s0y2.png" alt="New Component Code" width="800" height="944"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Finally, Copilot identified that it was necessary to register the new route in the main routes file and made this modification automatically:&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fovzqnai61djek0y0zju9.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fovzqnai61djek0y0zju9.png" alt="GitHub Copilot Chat Response" width="371" height="451"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;The new route was added along with the import of the newly created page:&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2F60q0t3zi2l0njukcjgst.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2F60q0t3zi2l0njukcjgst.png" alt="New Route Addition" width="800" height="864"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;All of this was generated from a single message in the Agent Mode chat. This is just one example of the countless possibilities this new feature offers. For now, Agent Mode is still in preview, but I have already noticed a significant increase in productivity while using it. I believe that in the future, it will become even more refined and a great ally in development, reducing rework and eliminating the need to copy code from other sources.&lt;/p&gt;

</description>
      <category>githubcopilot</category>
      <category>productivity</category>
      <category>ai</category>
      <category>programming</category>
    </item>
    <item>
      <title>Agent Mode do GitHub Copilot: Um Assistente de Código Ainda Mais Inteligente</title>
      <dc:creator>Guilherme Vilas Boas Roberto</dc:creator>
      <pubDate>Tue, 04 Mar 2025 18:22:49 +0000</pubDate>
      <link>https://forem.com/guintoki/agent-mode-do-github-copilot-um-assistente-de-codigo-ainda-mais-inteligente-3cme</link>
      <guid>https://forem.com/guintoki/agent-mode-do-github-copilot-um-assistente-de-codigo-ainda-mais-inteligente-3cme</guid>
      <description>&lt;p&gt;Recentemente, o Agent Mode do GitHub Copilot foi disponibilizado no preview do VSCode Insiders. Esse novo modo é capaz de iterar sobre seu próprio código, reconhecer erros e corrigi-los automaticamente. Além disso, pode sugerir comandos de terminal e solicitar sua execução, bem como analisar erros em tempo de execução e aplicar autocorreções.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fvmszzj9u6znwlx3n6p45.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fvmszzj9u6znwlx3n6p45.png" alt="Agent mode exemplo" width="382" height="389"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Agora, além de executar apenas a tarefa solicitada, o Copilot também pode adicionar tarefas adicionais que não foram especificadas no prompt, mas que são necessárias para que a solução funcione corretamente. Ele é capaz de identificar seus próprios erros e aplicar correções diretamente no código.&lt;/p&gt;

&lt;p&gt;Tenho testado o Agent Mode há algumas semanas em projetos experimentais para entender melhor seu funcionamento e como ele pode impactar minha produtividade no desenvolvimento.&lt;/p&gt;

&lt;h3&gt;
  
  
  Um exemplo prático
&lt;/h3&gt;

&lt;p&gt;Uma das funcionalidades mais interessantes do Agent Mode é sua capacidade de analisar um projeto inteiro (ou apenas arquivos específicos) e gerar soluções automaticamente. Ele pode até mesmo criar novos arquivos quando necessário.&lt;/p&gt;

&lt;p&gt;Para testar isso, criei um projeto simples de CRUD e solicitei ao Copilot que adicionasse uma nova rota para listar caminhões e motoristas em conjunto, combinando duas rotas já existentes:&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Frf9kr48md8zbg05osb0g.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Frf9kr48md8zbg05osb0g.png" alt="Solicitação copilot chat" width="375" height="128"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;O Agent Mode iniciou a análise do código, identificou referências relevantes dentro do projeto e determinou o caminho ideal para aplicar a solução solicitada.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fduhycwqlel8w39ph1vy2.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fduhycwqlel8w39ph1vy2.png" alt="Resposta do copilot" width="374" height="603"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Logo após, ele criou um novo arquivo, incluindo automaticamente os imports dos componentes necessários para a listagem e as requisições.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2F3yvdompw219te217kn8u.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2F3yvdompw219te217kn8u.png" alt="Resposta do Copilot chat" width="374" height="226"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;A seguir, adicionou a requisição para listar os caminhões e motoristas, incluindo a lógica de filtragem:&lt;br&gt;
&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fchxlnfltsxcyiutjo4qc.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fchxlnfltsxcyiutjo4qc.png" alt="Código do novo componente criado" width="800" height="699"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;E estruturou a exibição dos resultados:&lt;br&gt;
&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fepvhanfgyyxaohp8gfvc.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fepvhanfgyyxaohp8gfvc.png" alt="Código do novo componente criado" width="800" height="944"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Por fim, o Copilot identificou que era necessário registrar a nova rota no arquivo de rotas principal e fez essa modificação automaticamente:&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fkilhemx5j0d8ok07isav.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fkilhemx5j0d8ok07isav.png" alt="Resposta do chat do Github copilot" width="371" height="451"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;A nova rota foi adicionada junto com o import da página criada:&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fagkv496alheqf2423in2.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fagkv496alheqf2423in2.png" alt="Adição da nova rota" width="800" height="864"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Tudo isso foi gerado a partir de uma única mensagem no chat do Agent Mode. Esse é apenas um exemplo das inúmeras possibilidades dessa nova funcionalidade. Por enquanto, o Agent Mode ainda está em preview, mas já percebi um aumento significativo na produtividade ao utilizá-lo. Acredito que, no futuro, ele será ainda mais refinado e se tornará um grande aliado no desenvolvimento, reduzindo retrabalho e eliminando a necessidade de copiar código de outros lugares.&lt;/p&gt;

</description>
      <category>githubcopilot</category>
      <category>productivity</category>
      <category>ai</category>
      <category>github</category>
    </item>
    <item>
      <title>Aprendendo e Compartilhando: Minha Jornada com IA, Copilot e LLMs pt.1😁</title>
      <dc:creator>Guilherme Vilas Boas Roberto</dc:creator>
      <pubDate>Thu, 27 Feb 2025 00:34:18 +0000</pubDate>
      <link>https://forem.com/guintoki/aprendendo-e-compartilhando-minha-jornada-com-ia-copilot-e-llms-pt1-oj9</link>
      <guid>https://forem.com/guintoki/aprendendo-e-compartilhando-minha-jornada-com-ia-copilot-e-llms-pt1-oj9</guid>
      <description>&lt;p&gt;🚀 GitHub Copilot: Como Ele Transforma Seus Prompts em Código Inteligente 🤖💡&lt;br&gt;
Já se perguntou o que acontece nos bastidores quando o GitHub Copilot gera código? O processo não é mágica, há um fluxo estruturado que garante sugestões seguras, relevantes e funcionais. Vou te mostrar como isso funciona. 👇&lt;br&gt;
🔹 Fluxo de Entrada&lt;br&gt;
1️⃣ Contexto &amp;amp; Segurança 🔐&lt;br&gt;
O Copilot recebe seu prompt via HTTPS e analisa todo o ambiente ao redor: código antes e depois do cursor, tipo de arquivo, abas abertas, estrutura do projeto e frameworks em uso.&lt;br&gt;
2️⃣ Filtragem Inicial 🔍&lt;br&gt;
Antes de processar o prompt, o tráfego passa por um proxy para garantir integridade e impedir manipulações.&lt;br&gt;
3️⃣ Detecção de Conteúdo Sensível 🚫&lt;br&gt;
Um sistema de segurança bloqueia termos inadequados, discurso de ódio e dados pessoais antes da geração de qualquer sugestão.&lt;br&gt;
4️⃣ Geração de Código com IA ⚡&lt;br&gt;
O modelo de inteligência artificial interpreta o prompt e produz sugestões contextualizadas com base no código e no projeto.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fufxlaiqbgfapye713z9u.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fufxlaiqbgfapye713z9u.png" alt="Diagrama fluxo de entrada" width="800" height="656"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;🔹 Fluxo de Saída&lt;br&gt;
5️⃣ Validação &amp;amp; Segurança 🔎&lt;br&gt;
O código gerado passa por verificações contra vulnerabilidades como XSS e SQL Injection e pode ser comparado com repositórios públicos para evitar cópias diretas.&lt;br&gt;
6️⃣ Interação &amp;amp; Aprendizado 📈&lt;br&gt;
Você recebe as sugestões e, ao aceitá-las, rejeitá-las ou modificá-las, o Copilot aprende com seu estilo e melhora recomendações futuras.&lt;br&gt;
7️⃣ Aprimoramento Contínuo 🔁&lt;br&gt;
A cada novo prompt, o Copilot refina seu entendimento, tornando a experiência mais precisa e eficiente.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fep52shtjudsnnr5lvxei.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fep52shtjudsnnr5lvxei.png" alt="Diagrama fluxo de entrada" width="800" height="663"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;No fim das contas, o Copilot não é só um gerador de código, ele aprende com você e se adapta ao seu fluxo de trabalho. Quanto mais você usa, mais ele se ajusta às suas necessidades.&lt;/p&gt;

</description>
      <category>programming</category>
      <category>githubcopilot</category>
      <category>learning</category>
      <category>ai</category>
    </item>
  </channel>
</rss>
