<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Engineerageless on Medium]]></title>
        <description><![CDATA[Stories by Engineerageless on Medium]]></description>
        <link>https://medium.com/@engineerageless?source=rss-2b18fb5859fc------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*gK4srloE-Og39dwM</url>
            <title>Stories by Engineerageless on Medium</title>
            <link>https://medium.com/@engineerageless?source=rss-2b18fb5859fc------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 18:19:11 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@engineerageless/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[     ]]></title>
            <link>https://medium.com/@engineerageless/-b67cee643f37?source=rss-2b18fb5859fc------2</link>
            <guid isPermaLink="false">https://medium.com/p/b67cee643f37</guid>
            <dc:creator><![CDATA[Engineerageless]]></dc:creator>
            <pubDate>Wed, 21 Oct 2026 11:17:17 GMT</pubDate>
            <atom:updated>2026-10-21T11:17:17.009Z</atom:updated>
            <content:encoded><![CDATA[<p>Last night, close to midnight, I finally saw those green ticks in Azure DevOps.</p><p>After several failed deployments, SSH troubleshooting, and moments of “I am tired oooo,” my pipeline finally succeeded!</p><p>This was my first hands-on experience with Azure DevOps, and it helped me connect the dots between infrastructure, server configuration, source control, and automated application delivery.</p><p>What is Azure DevOps?</p><p>Azure DevOps is Microsoft’s platform for planning work, managing source code, and automating software delivery. Its services include Azure Boards, Azure Repos, Azure Pipelines, Azure Test Plans, and Azure Artifacts. For this project, I focused on Azure Repos for version control and Azure Pipelines for deployment automation.[learn.microsoft]</p><p>One interesting lesson: using Azure DevOps does not mean your application must run on Azure. Azure Pipelines supports deployment to different environments, including AWS, which was my deployment target.</p><p>What I built</p><p>I connected an Azure DevOps self-hosted agent to a YAML deployment pipeline that delivered a personalized static website to an AWS EC2 instance running Nginx.</p><p>The workflow:</p><p>Azure Repos → Azure Pipelines → Self-hosted agent → SSH → AWS EC2 → Nginx → Website in the browser.</p><p>Rather than treating these as separate tools, I began to understand how each one contributes to the delivery process.</p><p>🔹 Self-hosted agent foundation</p><p>I used a self-hosted Azure Pipelines agent running on an Ubuntu VM, organized under my SelfHostedPool agent pool.</p><p>That agent provided the execution environment for my pipeline jobs. This helped me understand that writing YAML is only one part of the process, you also need an available machine to execute the work.</p><p>🔹 Source control with Azure Repos</p><p>I imported the provided static website repository into Azure Repos, personalized the webpage, and stored my deployment configuration in azure-pipelines.yml.</p><p>My application code and deployment workflow were now version-controlled together.</p><p>🔹 Infrastructure with Terraform</p><p>I used Terraform to provision the AWS EC2 deployment environment, including the Ubuntu instance and supporting network access.</p><p>This separated infrastructure provisioning from application deployment: Terraform prepared the destination rather than transferring website files on every pipeline run.</p><p>🔹 Server configuration with Ansible</p><p>I used Ansible to prepare the target server by configuring Nginx, the website directory, SSH access, and the permissions needed for deployment.</p><p>An important detail was ensuring that the pipeline’s SSH user could update /var/www/html while Nginx could read and serve the files.</p><p>🔹 Deployment automation with Azure Pipelines</p><p>The deployment checks included confirming that index.html existed and checking the local Nginx HTTP response.</p><p>The biggest blocker was SSH authentication.</p><p>The pipeline repeatedly reported:</p><p>“Failed to connect to remote machine. Verify the SSH service connection details.”</p><p>That message alone did not tell me enough.</p><p>With temporary SSH debug logging enabled on the target server, I could see that the connection reached the machine and completed key exchange. The failure happened afterward: the client attempted password authentication while the server allowed public-key authentication.</p><p>That changed the troubleshooting direction.</p><p>Instead of opening more ports or enabling password authentication, I corrected the SSH service connection by uploading the intended private-key file into its Private Key field and saving the connection.</p><p>The next pipeline run succeeded. ✅ Logs turn guessing into diagnosis.</p><p>𝗠𝘆 𝗯𝗶𝗴𝗴𝗲𝘀𝘁 𝘁𝗮𝗸𝗲𝗮𝘄𝗮𝘆<br>Terraform, Ansible, and Azure Pipelines have different responsibilities, and work better when those responsibilities are clear.<br>A connection error is not automatically a network problem. Identifying the exact failure stage matters.<br>Credential configuration deserves careful attention. A masked field does not prove the correct credential is stored.<br>Deployment is not finished just because files were copied. Verification needs to be part of the workflow.<br>Security belongs in a learning project too. I used SSH private-key authentication and kept secrets out of the repository and pipeline YAML.<br>Cleaning up matters. After the lab, I destroyed the provisioned infrastructure rather than leaving the environment running unnecessarily.</p><p>This was not a smooth, straight-line exercise. It involved failed runs, troubleshooting, correcting assumptions, and trying again.</p><p>I finished with a clearer understanding of how code moves from a repository to a running website, and how to investigate when that journey breaks.</p><p>My first time with Azure DevOps was challenging, tiring, and absolutely worth it.</p><p>𝐇𝐚𝐯𝐞 𝐘𝐨𝐮 𝐭𝐫𝐢𝐞𝐝 𝐀𝐳𝐮𝐫𝐞 𝐃𝐞𝐯𝐎𝐩𝐬?</p><p>P.S. This post is part of the DevOps Micro Internship (DMI) with Agentic AI — Cohort 3 — by Pravin Mishra. My graded progress is public: <a href="https://dmi.pravinmishra.com/s/EAgeless.html">https://dmi.pravinmishra.com/s/EAgeless.html</a> · Start your DevOps journey: <a href="https://dmi.pravinmishra.com/?utm_source=student&amp;utm_medium=ps-blog&amp;utm_campaign=cohort3">https://dmi.pravinmishra.com/?utm_source=student&amp;utm_medium=ps-blog&amp;utm_campaign=cohort3</a></p><p>#AzureDevOps #AzurePipelines #AWS #EC2 #Terraform #Ansible #CICD #InfrastructureAsCode #DevOps #LearningInPublic</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*HAnhnJYkB5LSaDMR7J6ihA.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b67cee643f37" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[ANSIBLE FINALLY]]></title>
            <link>https://medium.com/@engineerageless/ansible-finally-2d09dc8d692f?source=rss-2b18fb5859fc------2</link>
            <guid isPermaLink="false">https://medium.com/p/2d09dc8d692f</guid>
            <dc:creator><![CDATA[Engineerageless]]></dc:creator>
            <pubDate>Fri, 09 Oct 2026 10:32:39 GMT</pubDate>
            <atom:updated>2026-10-09T10:32:39.502Z</atom:updated>
            <content:encoded><![CDATA[<p>I used to overlook Ansible. I thought that knowing Terraform and being able to provision cloud infrastructure was enough.</p><p>Terraform can create the foundation: resource groups, virtual networks, subnets, network security groups, public IP addresses, virtual machines, and databases. But infrastructure is only the beginning. A virtual machine can exist and still not be ready to serve an application.</p><p>Ansible has made me understand what happens after infrastructure is created.</p><p>It helps answer the operational questions that matter in the real world:</p><p>Is the server configured consistently?</p><p>Are the required packages installed?</p><p>Is Nginx configured correctly?</p><p>Is the application code deployed?</p><p>Are environment and database settings correct?</p><p>Are services enabled and running?</p><p>Can changes be tested safely before they are applied?</p><p>Can a team repeat the same configuration across multiple servers without manual mistakes?</p><p>In my recent EpicBook project, I used Ansible to manage an application server and validate that it was reachable, correctly configured, and capable of running the required deployment tasks. I also used Terraform to maintain the Azure infrastructure and update network access when my public IP address changed.</p><p>That experience made the relationship between Infrastructure as Code and configuration management much clearer to me:</p><p>Terraform creates the infrastructure.</p><p>Ansible configures and maintains what runs on that infrastructure.</p><p>I also built a small Ansible change-risk-review workflow. Instead of blindly running a playbook, the workflow first runs:</p><p>ansible-playbook — check — diff</p><p>This allows Ansible to show what it would change without actually changing the managed server.</p><p>The workflow then captures the output, checks the PLAY RECAP, identifies changed tasks, and flags categories such as:</p><p>Package-related changes</p><p>Service restart or reload changes</p><p>User and group changes</p><p>File permission or ownership changes</p><p>Configuration and template changes</p><p>Firewall and network-related changes</p><p>File removal or deletion changes</p><p>One controlled test involved creating a temporary file on my lab VM and adding an Ansible task that would remove it. Before applying the playbook, the dry-run report detected the proposed deletion and returned a warning. That may sound simple, but it reflects an important real-world engineering principle: even a small change should be understood before it is applied.</p><p>A command that removes a temporary file in a lab could remove a configuration file, an application artifact, a backup, or an important directory in a production environment if it is poorly reviewed.</p><p>This project helped me understand that automation is not only about speed. Good automation includes:</p><p>Evidence before action</p><p>Clear risk classification</p><p>Safe dry runs</p><p>Human review for impactful changes</p><p>Verification after deployment</p><p>Repeatable and documented processes</p><p>I also learned that an exit code does not always mean “success” or “failure” in the usual sense. In my review workflow:</p><p>Exit code 0 meant the environment was healthy.</p><p>Exit code 1 meant the dry run succeeded but a potentially risky change required review.</p><p>Exit code 2 meant the review itself failed, or Ansible could not validate the target environment.</p><p>That distinction is valuable because it turns automation into a decision-support system rather than a blind execution tool.</p><p>The final workflow followed a simple but powerful pattern:</p><p>Gather → Analyze → Human Act → Verify</p><p>Gather: Run Ansible in — check — diff mode and collect evidence.</p><p>Analyze: Review changed tasks, risk indicators, and the Ansible recap.</p><p>Human Act: Decide whether the proposed change is acceptable and manually apply it.</p><p>Verify: Confirm the playbook completed successfully, the host remains reachable, and the post-apply review reflects the expected state.</p><p>I am leaving this project with a stronger understanding of infrastructure operations. I can now provision with Terraform, configure with Ansible, preview changes before applying them, recognize risky actions, and verify outcomes after deployment.</p><p>Terraform gave me the ability to create infrastructure.</p><p>Ansible is teaching me how to operate it responsibly.</p><p>P.S. This post is part of the DevOps Micro Internship (DMI) with Agentic AI — Cohort 3 — by Pravin Mishra. My graded progress is public: <a href="https://dmi.pravinmishra.com/s/EAgeless.html">https://dmi.pravinmishra.com/s/EAgeless.html</a> · Start your DevOps journey: <a href="https://dmi.pravinmishra.com/?utm_source=student&amp;utm_medium=ps-blog&amp;utm_campaign=cohort3">https://dmi.pravinmishra.com/?utm_source=student&amp;utm_medium=ps-blog&amp;utm_campaign=cohort3</a></p><p>#Ansible #Terraform #DevOps #CloudEngineering #InfrastructureAsCode #Azure #Automation #ConfigurationManagement #Linux #Cybersecurity #SRE #ContinuousLearning</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*MdYuMdLjLZz43ztVk4Yfzg.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2d09dc8d692f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Week 8 has been all about turning cloud concepts into practical infrastructure.]]></title>
            <link>https://medium.com/@engineerageless/week-8-has-been-all-about-turning-cloud-concepts-into-practical-infrastructure-35c93f340ee6?source=rss-2b18fb5859fc------2</link>
            <guid isPermaLink="false">https://medium.com/p/35c93f340ee6</guid>
            <dc:creator><![CDATA[Engineerageless]]></dc:creator>
            <pubDate>Tue, 15 Sep 2026 08:48:26 GMT</pubDate>
            <atom:updated>2026-09-15T08:48:26.262Z</atom:updated>
            <content:encoded><![CDATA[<p>Through this week’s Terraform assignments, I have worked across Microsoft Azure and AWS to provision virtual machines, configure public access, deploy applications, connect cloud resources, and troubleshoot real deployment challenges.</p><p>The goal was not simply to create resources in the cloud. It was to understand how infrastructure is designed, automated, connected, and made usable.</p><p>Working through these projects has strengthened my practical experience with:</p><p>👉 Infrastructure as Code using Terraform.<br> 👉 Provisioning resources on Azure and AWS.<br> 👉 Configuring Linux virtual machines.<br> 👉 Managing SSH access and remote servers.<br> 👉 Deploying applications to cloud environments.<br> 👉 Connecting applications to managed databases.<br> 👉 Working with AWS EC2 and Amazon RDS.<br> 👉 Configuring MySQL databases and application data.<br> 👉 Using Nginx to serve applications.<br> 👉 Troubleshooting networking, ports, security rules, and application connectivity.</p><p>The process also taught me that cloud engineering requires more than following deployment commands. When an application is not accessible, a database is not connecting, or a server is showing the wrong page, every layer must be investigated, from the infrastructure and network configuration to the application and database.</p><p>Completing these assignments has helped me develop a more structured approach to cloud deployment: define the infrastructure, provision it consistently, configure the environment, test each connection, investigate failures, and verify the final result.</p><p>More importantly, this experience is building my confidence to work with real cloud environments and approach technical challenges with patience, curiosity, and a problem-solving mindset.</p><p>The most rewarding part has been seeing infrastructure created with code become a working, accessible application.</p><p>P.S. This post is part of the DevOps Micro Internship (DMI) with Agentic AI — Cohort 3 — by @Pravin Mishra. My graded progress is public: <a href="https://dmi.pravinmishra.com/s/EAgeless.html">https://dmi.pravinmishra.com/s/EAgeless.html</a> · Start your DevOps journey: <a href="https://dmi.pravinmishra.com/?utm_source=student&amp;utm_medium=ps-blog&amp;utm_campaign=cohort">https://dmi.pravinmishra.com/?utm_source=student&amp;utm_medium=ps-blog&amp;utm_campaign=cohort</a>3</p><p><a href="https://www.linkedin.com/search/results/all/?keywords=%23terraform&amp;origin=HASH_TAG_FROM_FEED">#Terraform</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23aws&amp;origin=HASH_TAG_FROM_FEED">#AWS</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23azure&amp;origin=HASH_TAG_FROM_FEED">#Azure</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23devops&amp;origin=HASH_TAG_FROM_FEED">#DevOps</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23cloudcomputing&amp;origin=HASH_TAG_FROM_FEED">#CloudComputing</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23infrastructureascode&amp;origin=HASH_TAG_FROM_FEED">#InfrastructureAsCode</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23cloudengineering&amp;origin=HASH_TAG_FROM_FEED">#CloudEngineering</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23linux&amp;origin=HASH_TAG_FROM_FEED">#Linux</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23mysql&amp;origin=HASH_TAG_FROM_FEED">#MySQL</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23nginx&amp;origin=HASH_TAG_FROM_FEED">#Nginx</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23continuouslearning&amp;origin=HASH_TAG_FROM_FEED">#ContinuousLearning</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/930/1*K7j-f57Y1IHp2WVh_2RevA.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=35c93f340ee6" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Microsoft Azure Finally — Week 07]]></title>
            <link>https://medium.com/@engineerageless/microsoft-azure-finally-week-07-2f9115b7a3b2?source=rss-2b18fb5859fc------2</link>
            <guid isPermaLink="false">https://medium.com/p/2f9115b7a3b2</guid>
            <dc:creator><![CDATA[Engineerageless]]></dc:creator>
            <pubDate>Fri, 21 Aug 2026 20:30:21 GMT</pubDate>
            <atom:updated>2026-08-21T20:30:21.378Z</atom:updated>
            <content:encoded><![CDATA[<p>I have spent years learning and working around DevOps and cloud concepts, but one platform I kept postponing was <strong>Microsoft Azure</strong>.</p><p>Not because Azure is not important; far from it. I simply got more comfortable with AWS, and when you are familiar with one cloud platform, it is easy to keep choosing what you already know.</p><p>But this week, the DevOps Micro-Internship pushed me to finally get hands-on with Azure. And I’m glad it did.</p><p>What I discovered is that learning Azure is not about forgetting AWS or starting DevOps from scratch. The concepts are still familiar: compute, storage, networking, identity, load balancing, databases, security, monitoring, and cost management. What changes is the platform’s language, interface, tooling, and ecosystem.</p><p>During the week, I worked with Azure Virtual Machines, Azure Storage, networking, load balancing, MySQL, and three-tier architecture. I deployed applications, configured access, worked with ports and security rules, hosted a static site, and explored how a multi-tier application should separate the presentation, application, and data layers.</p><p>I also had to troubleshoot. My VM deployment did not work immediately, and that was part of the learning. Azure forced me to look at VM sizes, regions, subscription constraints, validation errors, quotas, and configuration choices instead of just clicking “Create” repeatedly.</p><p>That was a useful reminder: cloud engineering is never only about provisioning a resource. It is about understanding <em>why</em> a resource is unavailable, how identity and subscription limits affect deployment, what security rules are required, and what the cost and operational impact will be.</p><p>One thing I appreciate more now is that Azure and AWS are not enemies competing for a winner. They both solve similar infrastructure problems, but they shine in different contexts. AWS has incredible breadth and a deeply mature cloud-native ecosystem. Azure is especially strong for enterprise environments that already run on Microsoft identity, productivity, Windows, and hybrid infrastructure.</p><p>The real advantage for a DevOps engineer is not loyalty to one cloud provider. It is being able to understand the architecture, ask the right questions, and choose the right platform for the organization and workload.</p><p>This week was my first serious hands-on step into Azure. I am still learning it, still troubleshooting it, and still getting used to its way of doing things — but that is exactly the point.</p><p>Growth is not always about adding another certificate or reading another documentation page. Sometimes it is about finally stepping into the platform you kept avoiding and learning it from the inside.</p><p><strong>Azure is no longer the cloud platform I keep postponing. It is now part of my multi-cloud journey.</strong></p><p>P.S. This post is part of the DevOps Micro Internship (DMI) with Agentic AI — Cohort 3 — by @<a href="https://www.linkedin.com/in/pravin-mishra-aws-trainer/">Pravin Mishr</a>a. My graded progress is public: https://dmi.pravinmishra.com/s/EAgeless.html · Start your DevOps journey: <a href="https://dmi.pravinmishra.com/?utm_source=student&amp;utm_medium=ps-blog&amp;utm_campaign=cohort3">https://dmi.pravinmishra.com/?utm_source=student&amp;utm_medium=ps-blog&amp;utm_campaign=cohort3</a></p><p>#Azure #AWS #DevOps #CloudEngineering #MultiCloud #MicrosoftAzure #AzureVM #CloudArchitecture #CloudSecurity #LearningInPublic #DMI #DevOpsJourney</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/447/1*iciyu9q9Yeh9CPC6SzYHlg.jpeg" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2f9115b7a3b2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Week 06 Was All About AWS]]></title>
            <link>https://medium.com/@engineerageless/week-06-was-all-about-aws-719e8ad1df40?source=rss-2b18fb5859fc------2</link>
            <guid isPermaLink="false">https://medium.com/p/719e8ad1df40</guid>
            <category><![CDATA[devops]]></category>
            <category><![CDATA[aws]]></category>
            <category><![CDATA[learning]]></category>
            <dc:creator><![CDATA[Engineerageless]]></dc:creator>
            <pubDate>Mon, 17 Aug 2026 04:50:46 GMT</pubDate>
            <atom:updated>2026-08-17T04:50:46.346Z</atom:updated>
            <content:encoded><![CDATA[<p>Week 06 Was All About AWS</p><p>Week 6 was one of the most practical stages of my DevOps journey so far. It was not about learning AWS services in isolation; it was about seeing how cloud resources work together to host applications, manage traffic, protect data, control spend, and deliver reliably.</p><p>Before deploying resources, I configured a monthly budget and cost alerts. AWS is usage-based, so cloud engineering includes financial responsibility from the start.</p><p>The Free Tier is useful for learning, but it doesn’t guarantee every resource is free. Public IPv4 addresses, NAT Gateways, extra EBS storage, RDS usage, and data transfer can create costs depending on how you configure them.</p><p>This week reinforced a simple operating principle: <strong>create intentionally, monitor continuously, and clean up what you no longer need.</strong></p><h3>From static hosting to full-stack deployment</h3><p>I deployed a personal portfolio website on <strong>Amazon S3</strong>, learning how to serve static sites without managing an operating system or web server.</p><p>I then deployed a mini finance website on <strong>Amazon EC2</strong>, where I took responsibility for the server itself: SSH access, Ubuntu packages, application files, ports, security groups, and verifying the app locally and externally.</p><p>The deeper deployment was <strong>EpicBook</strong>, which combined:</p><ul><li>EC2 and Ubuntu for compute</li><li>Node.js for the application</li><li>Nginx for web traffic</li><li>Amazon RDS for managed MySQL</li><li>Sequelize for database interaction</li><li>Security groups for controlled network access</li></ul><p>This was where troubleshooting became very real. I resolved stale Ubuntu package metadata that caused Nginx installation failures, corrected an invalid RDS endpoint that produced a DNS ENOTFOUND error, and learned that stopping a foreground Node.js process with Ctrl+C also stops the app listening on port 8080.</p><p>Those issues helped me understand the connection between application processes, ports, DNS, database connectivity, Linux process management, and network configuration.</p><p>I also worked with a highly available two-tier deployment and a three-tier Book Review capstone architecture.</p><p>The key shift was moving from:</p><p><em>“How do I deploy an application?” </em>to <em>“How do I design an application that remains available when infrastructure fails?”</em></p><p>High availability means reducing single points of failure through health checks, traffic distribution, scaling, network design, and resilient data services.</p><p>The three-tier architecture made separation of concerns clear:</p><ul><li><strong>Presentation tier:</strong> receives user requests.</li><li><strong>Application tier:</strong> runs backend logic.</li><li><strong>Data tier:</strong> stores persistent data securely.</li></ul><p>This approach improves scalability, maintainability, and security. The database should not be exposed directly to the public; the application tier should be the controlled bridge between users and data.</p><p>I finished the week by auditing AWS resources and investigating cost anomalies. This highlighted an important reality: deleting an EC2 instance does not always remove everything associated with it.</p><p>A responsible audit includes checking:</p><ul><li>EC2 instances</li><li>EBS volumes and snapshots</li><li>Elastic IP addresses</li><li>NAT Gateways</li><li>RDS databases</li><li>Security groups</li><li>Resources across regions</li><li>Billing line items and Cost Explorer usage types</li></ul><p>This is where DevOps meets <strong>FinOps</strong>. A deployment is not truly complete if it leaves unused resources running, creates unexpected costs, or lacks visibility.</p><h3>What’s next in AWS</h3><p>AWS is evolving beyond traditional infrastructure hosting. The direction is increasingly shaped by:</p><ul><li>AI-native and agentic workloads</li><li>Serverless architectures</li><li>Security automation and governance</li><li>Observability through logs, metrics, tracing, and alerts</li><li>FinOps and continuous cost optimization</li><li>Resilient, policy-driven cloud systems</li></ul><p>The engineer’s role is also evolving. Deploying an app successfully is no longer enough; we must understand its security posture, availability, performance, operational behavior, and cost impact.</p><h3>My Week 6 takeaway</h3><p>Week 6 strengthened my hands-on experience across S3, EC2, RDS, Nginx, Node.js, networking, security groups, high availability, three-tier architecture, troubleshooting, and cloud cost management.</p><p>AWS is not just a collection of services. It is an ecosystem for building and operating complete systems.</p><p><strong>Week 06 was all about AWS — but more importantly, it was about learning to think like a cloud engineer: design intentionally, deploy carefully, monitor continuously, secure every layer, and understand the operational and financial impact of every resource.</strong></p><p>P.S. This post is part of the DevOps Micro Internship (DMI) with Agentic AI — Cohort 3 — by Pravin Mishra. My graded progress is public: <a href="https://dmi.pravinmishra.com/s/EAgeless.html">https://dmi.pravinmishra.com/s/EAgeless.html</a> · Start your DevOps journey: <a href="https://dmi.pravinmishra.com/?utm_source=student&amp;utm_medium=ps-blog&amp;utm_campaign=cohort3">https://dmi.pravinmishra.com/?utm_source=student&amp;utm_medium=ps-blog&amp;utm_campaign=cohort3</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*51wvTkQzpqkSFRI9QjZNPw.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=719e8ad1df40" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Week 5 Stronger and Better]]></title>
            <link>https://medium.com/@engineerageless/week-5-stronger-and-better-cc4a5ab156ca?source=rss-2b18fb5859fc------2</link>
            <guid isPermaLink="false">https://medium.com/p/cc4a5ab156ca</guid>
            <category><![CDATA[engineering]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[devops]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[growth]]></category>
            <dc:creator><![CDATA[Engineerageless]]></dc:creator>
            <pubDate>Tue, 11 Aug 2026 12:32:37 GMT</pubDate>
            <atom:updated>2026-08-17T05:12:18.821Z</atom:updated>
            <content:encoded><![CDATA[<p>Week 5 was the point where my Jira board stopped being “just a place to move cards” and became a data source I could actively analyze with AI.</p><p>Across the assignments this week, I did three big things:</p><ol><li><strong>Got my Jira foundation in place</strong><br>I created my Jira account, set up my profile, and built a Scrum board for the DevOps Micro‑Internship website. I defined a backlog, created a sprint with multiple user stories, and made sure each story had a clear summary, type, status, assignee, and estimate. This is the basic structure you need before you can talk seriously about sprint health.</li></ol><p><strong>2. Practiced running and shipping a real mini sprint</strong><br>I didn’t stay at the “toy example” level. I ran a 5‑day mini sprint, treated the work as an increment, and focused on shipping something visible. That meant using Jira not just for tracking, but for planning, refining, and closing work in a way that matches how teams actually operate.</p><p><strong>3. Connected AI directly to my Jira board using MCP</strong><br>The real upgrade came with the AI‑assisted sprint health assignment. I wired Claude Code to my live Jira board through an MCP server. Technically, that meant:</p><ul><li>Creating a Jira API token, tied to my Atlassian account.</li><li>Configuring a jira MCP server in .mcp.json using uvx mcp-atlassian.</li><li>Storing my Jira URL, username, and API token in .claude/settings.local.json, which is gitignored so my secrets never go into the repo.</li><li>Verifying the connection with /mcp until Jira showed as <strong>connected</strong>, not just configured.</li></ul><p>Once that was working, I asked Claude to read my <strong>active sprint</strong> through Jira’s read‑only API and return real data: issue keys, statuses, story points, assignees, and priorities. I cross‑checked the response against my board in the browser to make sure it wasn’t guessing. The lists matched, which proved I was looking at live sprint data, not model hallucinations.</p><h3>What “AI‑assisted sprint health” actually means</h3><p>Using AI here is not about letting it “run the sprint for me.” It’s about using it to <strong>analyze the sprint in a structured way</strong> so I can make better decisions faster.</p><p>I turned that one‑off query into a reusable skill called /sprint-health. In plain terms, here’s what it does:</p><ul><li>It <strong>locates the current active sprint</strong> on my Scrum board via Jira MCP.</li><li>It <strong>pulls every issue in that sprint</strong>, including status, story points, assignee, and last updated timestamp.</li><li>It then calculates:</li><li>How many story points are already in “Done” versus how many were committed at the start.</li><li>How many days remain in the sprint.</li><li>Which stories are at risk (for example, still “To Do” or “In Progress” with very little time left, or not updated in several days).</li><li>Which items have missing estimates or weak descriptions (no story points, no clear acceptance criteria).</li></ul><p>The output is a <strong>sprint health report</strong>: a clear summary of velocity so far, items at risk of missing the sprint, and gaps in the backlog or sprint planning. Instead of me scanning every ticket and doing all this in my head, I get a focused view of where attention is needed.</p><h3>Why I kept the AI strictly read‑only</h3><p>One of the most important design choices I made was to keep the skill <strong>read‑only</strong>:</p><ul><li>In the skill’s definition, the allowed-tools list only includes Jira tools that read data (search, get issue, get sprint, get board) plus Read.</li><li>I deliberately <strong>excluded</strong> any tools that can create issues, update fields, transition statuses, or add comments.</li></ul><p>So even though Claude is connected to my Jira board, it cannot quietly move a story to “Done”, change the assignee, or edit the description. All of that remains in my hands.</p><p>This matters for project management because the board is the team’s source of truth. Sprint commitments, burndown charts, and stakeholder updates are all based on what the board says. If an AI could mutate the board without the Scrum Master noticing, you could easily end up with misleading data and broken trust.</p><p>By keeping AI in a read‑only role, I get the best of both worlds:</p><ul><li>Automated <strong>Gather</strong> and <strong>Analyze</strong> on top of real board data.</li><li>Human <strong>Act</strong> and <strong>Verify</strong> when it comes to actually changing tickets.</li></ul><p>When I want to adjust the sprint, I go into Jira myself, move a ticket or add an estimate, and then rerun /sprint-health to confirm the report reflects the new state. The skill proves the change happened, but it never performs the change.</p><h3>Why this feels “next level” for my engineering practice</h3><p>At a surface level, this week taught me how to plug Claude into Jira. But at a deeper level, it taught me how to think about AI as part of the delivery pipeline:</p><ul><li><strong>Jira</strong> holds my structured work data.</li><li><strong>MCP</strong> is the controlled bridge between Jira and Claude.</li><li><strong>Claude</strong> becomes an analytical layer over the sprint, not an untrusted actor inside it.</li></ul><p>Instead of using AI just for code suggestions or generic answers, I’m now using it to read my actual project state, compute real sprint metrics, and flag risk in a way that directly supports standups and planning.</p><p>For me, that’s what “Week 5 Stronger and Better” really means: I’m not only more comfortable with Jira and Scrum, I’m more deliberate about how I use AI inside project management, where it should observe and analyze, and where humans must still own the decisions and the board.</p><p>P.S. This post is part of the DevOps Micro Internship (DMI) with Agentic AI — Cohort 3 — by Pravin Mishra. My graded progress is public: <a href="https://dmi.pravinmishra.com/s/EAgeless.html">https://dmi.pravinmishra.com/s/EAgeless.html</a> · Start your DevOps journey: <a href="https://dmi.pravinmishra.com/?utm_source=student&amp;utm_medium=ps-blog&amp;utm_campaign=cohort3">https://dmi.pravinmishra.com/?utm_source=student&amp;utm_medium=ps-blog&amp;utm_campaign=cohort3</a></p><p>#Week5StrongAndBetter #DevOps #Jira #Scrum #MCP #ClaudeCode #AgenticAI #LearningInPublic #DMI #DevOpsJourney</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3QghsYNAMdj_-YPuK2TbaQ.jpeg" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cc4a5ab156ca" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Week 4 of the DevOps Micro Internship stretched me in the best way.]]></title>
            <link>https://medium.com/@engineerageless/week-4-of-the-devops-micro-internship-stretched-me-in-the-best-way-77e7f9456e01?source=rss-2b18fb5859fc------2</link>
            <guid isPermaLink="false">https://medium.com/p/77e7f9456e01</guid>
            <category><![CDATA[github]]></category>
            <category><![CDATA[devops]]></category>
            <category><![CDATA[git]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[internships]]></category>
            <dc:creator><![CDATA[Engineerageless]]></dc:creator>
            <pubDate>Fri, 24 Jul 2026 15:03:46 GMT</pubDate>
            <atom:updated>2026-07-24T15:03:46.405Z</atom:updated>
            <content:encoded><![CDATA[<p>This week, I revisited the foundations of Git and GitHub, branching, pull requests, forks, remotes, and the discipline of working on a feature branch instead of main. Then I went deeper with something more real: building a Git pre-commit hook that blocks secret-like patterns and oversized files before they ever become a commit.</p><p>The most interesting part was pairing that fixed-rule hook with a Claude Code /pr-ready skill. One checks for exact risks, the other reads the staged diff like a reviewer and flags context problems a rule-based check can miss. That contrast taught me something important: in DevOps, not every safety net should do the same job. Some should enforce, others should interpret, and a human should still make the final call.</p><p>What stood out to me most was how easy it is to feel “done” once automation catches something. But real engineering is not just catching mistakes; it is understanding why they happened, fixing the workflow, and making the next failure less likely. That is the kind of discipline I want to keep building.</p><p>I’m leaving this week with stronger Git habits, a better understanding of safety checks, and a clearer respect for the difference between AI assistance and human accountability.</p><p>What about you, what did you learn this week??</p><p>P.S. This post is part of the DevOps Micro Internship (DMI) with Agentic AI — Cohort 3 — by <a href="https://www.linkedin.com/in/pravin-mishra-aws-trainer/">Pravin Mishra</a>. My graded progress is public: <a href="https://dmi.pravinmishra.com/s/YOUR-GITHUB-USERNAME.html">https://dmi.pravinmishra.com/s/EAgeless.html</a> · Start your DevOps journey: <a href="https://dmi.pravinmishra.com/?utm_source=student&amp;utm_medium=ps-blog&amp;utm_campaign=cohort3">https://dmi.pravinmishra.com/?utm_source=student&amp;utm_medium=ps-blog&amp;utm_campaign=cohort3</a></p><p>#DevOps #Git #GitHub #ClaudeCode #AgenticAI #GitHooks #PullRequest #ContinuousImprovement #LearningInPublic #DevOpsJourney #AIinDevOps #InternshipJourney</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*S5_fTJlZlnFUjHkquty-DA.jpeg" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=77e7f9456e01" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[DevOps Internship]]></title>
            <link>https://medium.com/@engineerageless/devops-internship-6e56fa611d9a?source=rss-2b18fb5859fc------2</link>
            <guid isPermaLink="false">https://medium.com/p/6e56fa611d9a</guid>
            <dc:creator><![CDATA[Engineerageless]]></dc:creator>
            <pubDate>Wed, 22 Jul 2026 10:16:55 GMT</pubDate>
            <atom:updated>2026-07-22T10:16:55.495Z</atom:updated>
            <content:encoded><![CDATA[<p>I’m excited to share that I have officially started this journey with DMI Cohort 3.</p><p>This is the beginning of a 5-month learning experience that I’m truly grateful to be part of. I’m looking forward to building real-world projects, sharpening my DevOps skills, learning with Agentic AI, and growing through the support of mentors and the community.</p><p>What excites me most is not just the technical growth, but the discipline, mindset, and consistency this journey will require. I’m expecting to challenge myself, learn from every assignment, present my work publicly, and become more confident in the process.</p><p>I know this won’t just be about learning tools, it will be about becoming better, week by week, project by project, and learning how to show up with intention and purpose.</p><p>Thank you, <a href="https://www.linkedin.com/in/pravin-mishra-aws-trainer/">Pravin Mishra</a>, for the opportunity and for making this journey possible. I truly appreciate it.</p><p><a href="https://www.linkedin.com/search/results/all/?keywords=%23dmi&amp;origin=HASH_TAG_FROM_FEED">#DMI</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23dmicohort3&amp;origin=HASH_TAG_FROM_FEED">#DMIcohort3</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23devops&amp;origin=HASH_TAG_FROM_FEED">#DevOps</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23agenticai&amp;origin=HASH_TAG_FROM_FEED">#AgenticAI</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23learningjourney&amp;origin=HASH_TAG_FROM_FEED">#LearningJourney</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23growthmindset&amp;origin=HASH_TAG_FROM_FEED">#GrowthMindset</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23internship&amp;origin=HASH_TAG_FROM_FEED">#Internship</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23community&amp;origin=HASH_TAG_FROM_FEED">#Community</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23pravinmishra&amp;origin=HASH_TAG_FROM_FEED">#PravinMishra</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/480/0*Fp_EsxBfkm0Sb8W_" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6e56fa611d9a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Week 3 Done and Dusted]]></title>
            <link>https://medium.com/@engineerageless/week-3-done-and-dusted-9e9f062a8e62?source=rss-2b18fb5859fc------2</link>
            <guid isPermaLink="false">https://medium.com/p/9e9f062a8e62</guid>
            <category><![CDATA[dmibypravinmishra]]></category>
            <category><![CDATA[devops]]></category>
            <category><![CDATA[bash-scripting]]></category>
            <category><![CDATA[linux]]></category>
            <category><![CDATA[agentic-ai]]></category>
            <dc:creator><![CDATA[Engineerageless]]></dc:creator>
            <pubDate>Wed, 22 Jul 2026 10:11:59 GMT</pubDate>
            <atom:updated>2026-07-22T10:11:59.264Z</atom:updated>
            <content:encoded><![CDATA[<p>Week 3 was all about Bash scripting, and I’m genuinely proud of how much I built and learned along the way.</p><p>I started by getting comfortable with the basics of Bash and Linux command-line work, then moved into writing a script that could collect system information in a structured and repeatable way. From there, I built more advanced logic using variables, arrays, loops, conditionals, functions, command substitution, and exit codes so the script could do real work instead of just printing commands. I also learned how to validate scripts properly, make them executable, and use the right checks to confirm they run cleanly before testing them on a server.</p><p>One of the biggest lessons from this week was that Bash is not just about running commands — it is about creating reliable, reusable automation. I built a read-only Linux health check workflow that gathered evidence about service status, listening ports, local HTTP response, disk usage, memory, and service logs, then used that output to support incident analysis. That exercise taught me how to separate evidence collection from analysis and human-approved recovery, which is exactly the kind of discipline real DevOps work requires.</p><p>I also explored how agentic AI fits into this process by connecting the Bash report to Claude Code through a controlled, read-only workflow. Instead of letting the AI take action directly, I used it to analyze the collected evidence, explain the likely cause, and recommend a safe next step for human review. That made the workflow feel much more practical and realistic: Bash gathered the facts, AI interpreted them, and I remained the operator responsible for the final action.</p><p>By the end of the week, I had completed the full assignment workflow: building the triage script, documenting the project rules, creating the skill, simulating an Nginx incident, capturing failed-state evidence, recovering the service manually, and verifying everything again after recovery. That last verification step really reinforced the importance of not guessing during incidents — always gather evidence, analyze it carefully, act intentionally, and then confirm the system is healthy again.</p><p>P.S. This post is part of the DevOps Micro Internship (DMI) with Agentic AI — Cohort 3 — by @<a href="https://www.linkedin.com/in/pravin-mishra-aws-trainer/">Pravin Mishr</a>a. My graded progress is public: <a href="https://dmi.pravinmishra.com/s/YOUR-GITHUB-USERNAME.html">https://dmi.pravinmishra.com/s/EAgeless.html</a> · Start your DevOps journey: <a href="https://dmi.pravinmishra.com/?utm_source=student&amp;utm_medium=ps-blog&amp;utm_campaign=cohort3">https://dmi.pravinmishra.com/?utm_source=student&amp;utm_medium=ps-blog&amp;utm_campaign=cohort3</a></p><p><a href="https://www.linkedin.com/search/results/all/?keywords=%23dmibypravinmishra&amp;origin=HASH_TAG_FROM_FEED">#DMIByPravinMishra</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23bashscripting&amp;origin=HASH_TAG_FROM_FEED">#BashScripting</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23linux&amp;origin=HASH_TAG_FROM_FEED">#Linux</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23devops&amp;origin=HASH_TAG_FROM_FEED">#DevOps</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23agenticai&amp;origin=HASH_TAG_FROM_FEED">#AgenticAI</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23claudecode&amp;origin=HASH_TAG_FROM_FEED">#ClaudeCode</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23nginx&amp;origin=HASH_TAG_FROM_FEED">#Nginx</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23systemadministration&amp;origin=HASH_TAG_FROM_FEED">#SystemAdministration</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23automation&amp;origin=HASH_TAG_FROM_FEED">#Automation</a> <a href="https://www.linkedin.com/search/results/all/?keywords=%23learninginpublic&amp;origin=HASH_TAG_FROM_FEED">#LearningInPublic</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/896/1*1A4EuG6B4PD9nZCOZUz0tg.jpeg" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9e9f062a8e62" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Reflection, Week 2 —by Evangeline Obeta]]></title>
            <link>https://medium.com/@engineerageless/reflection-week-2-by-evangeline-obeta-090af287a4e3?source=rss-2b18fb5859fc------2</link>
            <guid isPermaLink="false">https://medium.com/p/090af287a4e3</guid>
            <category><![CDATA[dmibypravinmishra]]></category>
            <category><![CDATA[devops]]></category>
            <category><![CDATA[agentic-ai]]></category>
            <category><![CDATA[claude-code]]></category>
            <dc:creator><![CDATA[Engineerageless]]></dc:creator>
            <pubDate>Fri, 10 Jul 2026 22:18:11 GMT</pubDate>
            <atom:updated>2026-07-10T22:18:11.539Z</atom:updated>
            <content:encoded><![CDATA[<p>Week 2 stretched me in ways I did not expect. I thought I was just going to complete tasks around Claude Code, skills, subagents, MCP, hooks, permissions, and memory, but the week became a real lesson in patience, attention to detail, and how easy it is to go in the wrong direction when instructions are not followed carefully. This was one of those weeks where I had to face not just technical tasks, but also myself — my panic, my impatience, my assumptions, and my ability to recover when things did not go as planned.</p><h3>1. Biggest technical insight I got this week</h3><p>My biggest technical insight this week was understanding that Agentic AI is only powerful when there is structure around it. Before this week, it was easy to think of AI as something you just prompt and then wait for magic. But working through the Week 2 assignments showed me that Claude Code works best when its behavior is guided properly.</p><p>For example, I began to understand why CLAUDE.md matters so much. It is not just another file inside a project; it helps shape how the AI behaves inside that environment. I also saw how skills can turn repeated instructions into reusable workflows instead of rewriting the same long prompts again and again. Subagents made even more sense to me because they showed that work can be broken into smaller responsibilities instead of trying to make one assistant do everything at once.</p><p>Another important insight came from MCP, hooks, permissions, and memory. MCP made me see that AI can connect to tools and systems in a more practical way, but that power also comes with the need for control. Hooks and permissions showed me that not every tool should have unrestricted access, and memory helped me understand how consistency can be built into an AI workflow over time. This week helped me move from seeing AI as a chatbot to seeing it as a workflow partner that still needs rules, boundaries, and human direction.</p><h3>2. Biggest insight I got about myself this week</h3><p>The biggest thing I learned about myself this week is that I can be too eager to move quickly, especially when I feel pressure to finish. I noticed that once I believe I understand a task, I sometimes rush into execution before slowing down to confirm the exact instruction, folder structure, or expected outcome.</p><p>This week exposed that habit clearly. There were moments when I followed the wrong steps, created things in the wrong location, misunderstood what the structure was supposed to look like, and then spent extra time trying to fix what could have been avoided. That was frustrating, but it also taught me something important: I do not actually learn best when I rush. I learn better when I read carefully, test one step at a time, and pay attention to what the terminal or file structure is showing me.</p><p>I also realized that I am becoming more confident with the terminal than I used to be. Even though I panicked at some points, I still stayed with the work, asked questions, retried commands, and kept troubleshooting. That matters to me because it shows growth. I may not know everything yet, but I am getting more comfortable working through technical confusion instead of running away from it.</p><h3>3. My biggest weakness or loop I noticed</h3><p>My biggest weakness this week was skipping verification and then panicking when the result was not what I expected.</p><p>I saw this pattern repeat itself in different ways. Sometimes I assumed I was in the right folder and ran a command without checking. Sometimes I thought a structure was wrong when it was actually just a display issue. Sometimes I became so focused on finishing quickly that I missed small details that later turned into bigger blockers. And once the errors started showing up, panic would set in because submission was getting closer and I felt behind.</p><p>The hardest part was realizing that some of my stress came from avoidable mistakes. That was not easy to admit, but it was necessary. This week taught me that technical problems are one thing, but emotional pressure can make small problems feel much bigger. If I do not slow down and verify each step, I create a loop for myself: rush, miss detail, get confused, panic, restart. That is the loop I want to break.</p><h3>4. One system I will implement from this week</h3><p>The system I will implement going forward is simple: before running any command for an assignment, I will pause and verify three things — my current folder, the expected output, and whether the instruction says to create, edit, commit, or merge.</p><p>I will do this at the beginning of every study session and before every major command I run in the terminal. This habit will help me reduce avoidable mistakes, especially the kind that happen because I assumed I was in the right place or thought I already understood the step. It will also help me stay calmer, because instead of rushing and then troubleshooting under pressure, I will be working from a checklist mindset.</p><p>In practical terms, my system for next week will be:</p><ul><li>Read the instruction fully before touching the terminal.</li><li>Run pwd before commands that create or move files.</li><li>Compare my folder structure with the expected structure before continuing.</li><li>Check git status before and after important Git actions.</li><li>Take screenshots at the correct moment instead of promising myself I will do it later.</li></ul><p>This system may feel small, but I believe it will save me a lot of time and reduce unnecessary panic.</p><h3>5. What I learned about Agentic AI and DevOps</h3><p>This week changed how I think about both Agentic AI and DevOps. I used to think of AI mostly as something you talk to, but now I understand that Agentic AI is much more about systems, workflows, control, and repeatability. It is not just conversation — it is configuration, task design, permissions, context, and safe execution.</p><p>That connects strongly to DevOps. DevOps is not just about tools either; it is about creating reliable systems that people can trust. The same way infrastructure needs guardrails, AI workflows also need guardrails. The same way DevOps engineers care about automation but still review outputs carefully, Agentic AI also needs human oversight. This week showed me that automation is powerful, but unmanaged automation can create confusion very quickly.</p><p>I also learned that reusable skills reduce manual repetition, subagents can divide responsibility in a smart way, and permissions matter because safety matters. One of the clearest lessons for me is that human review is still necessary. AI can support the workflow, but it should not replace understanding. If anything, this week taught me that DevOps with AI requires even more intentional thinking, not less.</p><h3>6. My Week 2 highlight</h3><p>My biggest highlight this week was not a perfect moment — it was the moment I kept going even after several things went wrong.</p><p>I had moments of failure, moments of confusion, and moments where I realized I had followed the wrong path and had to start again. There was also real panic because the submission deadline was close and I still was not done. But I am proud that I did not stop. I kept asking questions, kept checking the outputs, kept fixing my structure, and kept moving until I understood more than I did at the start.</p><p>If I had to choose one moment I am proud of, it would be the point where I started understanding the difference between just doing commands and actually understanding why I was doing them. That shift mattered to me. By the end of the week, I was not just trying to survive the tasks — I was beginning to see the logic behind Claude Code, skills, subagents, MCP, hooks, permissions, and memory.</p><h3>Closing reflection</h3><p>Week 2 was tasking, messy, and honestly humbling, but it was also deeply valuable. It showed me that learning Agentic AI is not just about completing assignments; it is about building the discipline to read carefully, troubleshoot calmly, document honestly, and improve my workflow over time.</p><p>This week reminded me that growth in tech does not always look neat. Sometimes it looks like failure, correction, restarting, and pushing through panic. But even in all of that, there is progress. And for me, that is the real win from Week 2.</p><p>Thank you for Reading</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Wj-xcZcOQ7LH1Vlg-7L6Pw.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=090af287a4e3" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>