Forem: Cristian Sarmiento

Anthropic Just Cut Off Third-Party Tools from Claude Subscriptions — My No-BS Take as a Libertarian AI Engineer

Cristian Sarmiento — Fri, 10 Apr 2026 20:35:16 +0000

Hey folks,
I’m Cristian Sarmiento — AI Software Engineer, ex-satellite operations specialist at ARSAT (Argentina’s national space agency), autistic + ADHD, libertarian, and sometimes a horrorsynth composer when I need to step away from the code.

Like many of you, I just learned that Anthropic has pulled the plug on third-party tools like OpenClaw for Claude Pro and Max subscribers. Starting April 4th, 2026 at 12pm PT, these tools no longer work under your subscription. Period.

Boris Cherny (Head of Claude Code) explained it clearly: the subscription plans were never built for the extreme usage patterns these third-party agents generate — non-stop 24/7 requests, zero context reuse, and massive compute consumption.
Am I annoyed? Absolutely. I’ve built several agents that relied on Claude and this forces me to recalculate my costs. But surprised? Not at all.

I’m a libertarian, so let me be direct: a private company has every right to protect its infrastructure and resources. If third-party agents are hammering their GPUs around the clock, Anthropic must choose — either raise prices for everyone or implement limits. Choosing limits is the smarter move. It’s not anti-innovation. It’s responsible business.

What I find funny is seeing people who usually criticize capitalism getting upset when a company finally decides to charge based on actual usage. Welcome to how markets work.

For those of us building stuff: yes, this stings. You now have to pay via “Extra Usage” bundles (with some discount) or use your own API key. They did give a one-month credit as a buffer, which is decent.

My practical advice as an engineer who worked in satellite operations (where compute capacity was also extremely expensive):
• Properly measure the real consumption of your agents.
• Evaluate whether the new cost makes sense or if open-source alternatives are good enough.
• Diversify. Never put all your eggs in one AI provider’s basket.|

This situation is a healthy reminder that “unlimited cheap AI” was never truly unlimited. Those GPUs run on real electricity and real money.

What’s your plan? Did this change affect you directly? Are you switching providers, moving to raw API, or something else? Drop your thoughts below — I read every comment (and reply when I’m not deep in hyperfocus).

Abrazo from Córdoba, Argentina (mate in one hand, cat on the keyboard),
Cristian Sarmiento 🍸🍋
https://x.com/crisesarmiento

Google Finally Admits the Browser Was Broken

Cristian Sarmiento — Wed, 08 Apr 2026 00:04:35 +0000

Vertical tabs, reading mode, and the quiet war for your attention

I had 47 tabs open when I wrote this.

Not because I'm disorganized. Because I'm the kind of person whose brain doesn't let go of context easily. Every tab is a thought I'm not done with yet. Close it and the thought is gone. Keep it open and the tab bar becomes a horizontal strip of anxiety — 47 identical rectangles, each showing maybe 8 characters of text, none of them enough to tell you what's actually there.

This is not a personal problem. This is a design failure that shipped with every major browser for 20 years.

Chrome just acknowledged it.

What actually changed

Google shipped two new features this week: vertical tabs and immersive reading mode.

Vertical tabs move the tab strip to the side of the window. You see full page titles. You can organize by group. When you hit double digits, you don't lose them — you just scroll. You can collapse the sidebar to a column of favicons when you need the space back.

Immersive reading mode does something more radical: it takes a webpage and removes everything that isn't the content. No ads. No sidebars. No related articles fighting for your peripheral vision. Just text, properly typeset, full screen.

Two features. Both small in implementation. Both significant in what they admit.

The ADHD tax on the modern web

Context switching has a cost. Neuroscience has known this for decades — every time you shift attention, your brain pays a switching tax. For most people it's a minor overhead. For people with ADHD or autism, it compounds.

The horizontal tab bar was designed for a world where you had 5 tabs open. It has not been redesigned for the world where a working session involves 30. When you can't read tab titles, you can't navigate without clicking. When you can't navigate without clicking, every search for the right tab is an interruption. Interruptions break flow. Broken flow in a neurodivergent brain can cost 20 minutes of recovery time.

That's the ADHD tax. Paid in productivity, paid in frustration, paid thousands of times a year, invisible to anyone who hasn't felt it.

Vertical tabs eliminate that specific tax. Full titles, scannable in a single glance, without clicking anything. It sounds trivial until you realize how much cognitive load it silently offloads.

Reading mode goes further. The modern webpage is an adversarial environment by design — every element competes for your attention. For someone who struggles to filter irrelevant stimuli, reading anything online is an exercise in selective suppression. You're not just reading the article; you're actively ignoring everything around it.

Immersive reading mode does the suppression for you. One click, and the page becomes what it was supposed to be: something worth reading.

The part that should make Google uncomfortable

Here's where it gets interesting.

Google's core business is advertising. Chrome exists, in no small part, to keep users browsing — and browsing means seeing ads. Immersive reading mode strips ads. Google just shipped a feature that removes its own revenue from the equation.

That's not philanthropy. That's a strategic calculation.

The browser war is real. Arc built its entire identity on rethinking what a browser could be — tab management, spaces, split views, a UI that actually respected how knowledge workers think. Vivaldi went further with customization. Firefox pushed reader view years before Chrome considered it. Edge shipped vertical tabs in 2021.

Google watched market share move toward these browsers among the exact demographic that matters most: power users, developers, people who live in their browser for 10 hours a day. These users don't click ads anyway. But they influence purchasing decisions, they write about tools they use, and they leave when something better exists.

Keeping them in Chrome is worth more than the ad impressions they skip.

So Google ships reading mode — not despite the fact that it removes ads, but as a calculated trade. User retention beats marginal ad revenue from users who were tuning it out anyway.

That's the attention economy eating itself. And honestly? It's the right outcome.

What this validates

Arc figured this out in 2022. The indie browser builders understood something Google took four more years to ship: people don't want more features. They want less friction.

The best productivity tool is the one that gets out of your way.

Vertical tabs get out of your way. Reading mode gets out of your way. These aren't innovations — they're corrections. They're Google admitting, through a product update, that the default browser experience was costing users something real.

I'll take the correction.

The techno-optimist case here isn't that Chrome became the best browser overnight. It's that even the largest, most ad-dependent platform on earth can feel competitive pressure from better design — and respond. Slowly, yes. Late, definitely. But the direction is right.

Tools that respect how human brains actually work will win. Every time.

It just sometimes takes a decade for the incumbents to notice.

Follow me on X: @crisesarmiento

Sources:

The First Real Counterattack

Cristian Sarmiento — Tue, 07 Apr 2026 22:31:39 +0000

How Project Glasswing flips the AI security equation — and why it matters for every engineer alive

A 27-year-old bug was sitting in OpenBSD.

Not theoretical. Not a minor edge case. A high-severity zero-day, invisible to every security audit, every static analyzer, every fuzzer that had ever touched that codebase. For 27 years, it waited.

Claude Mythos found it in a matter of hours.

That single fact is all you need to understand why Anthropic just launched Project Glasswing — and why I think it's one of the most important things that happened in tech this year.

The problem nobody wanted to say out loud

For the last two years, the cybersecurity industry has been tiptoeing around an uncomfortable truth: AI is already better than almost any human at finding and exploiting vulnerabilities.

Not eventually. Now.

The window between vulnerability discovery and exploitation has collapsed. What used to take a skilled attacker months — reconnaissance, fuzzing, exploit development, chaining bugs — can now happen in minutes with the right model. CrowdStrike's CTO said it plainly at the Glasswing launch: "capabilities have crossed a threshold that fundamentally changes the urgency required to protect critical infrastructure."

We've known this was coming. We didn't have a plan.

The default response from the industry was to quietly not talk about it, keep building, and hope defenders would keep pace. They weren't keeping pace. Cyber attack costs are running around $500 billion a year globally, and that number was accelerating before frontier AI models entered the picture.

Project Glasswing is the first serious response.

What Mythos Preview actually is

Claude Mythos Preview is not a product. Anthropic has no plans to release it publicly — and the reason they give is direct: its cybersecurity capabilities make it too dangerous for general availability.

That's a sentence worth sitting with.

Anthropic built a model so capable at finding and understanding software vulnerabilities that they made a deliberate decision to keep it out of reach. Not because of regulatory pressure. Not because of PR. Because they ran the numbers and the asymmetry between attack and defense was too severe.

The benchmark tells the story: Mythos Preview scores 83.1% on CyberGym vulnerability reproduction tests. Claude Opus 4.6 — already one of the best models available — scores 66.6%. That's not an incremental improvement. That's a different category of capability.

In practice, it means:

A 27-year-old OpenBSD vulnerability that survived decades of audits
A 16-year-old FFmpeg bug that automated fuzzers hit 5 million times without catching
Autonomous exploit chains in the Linux kernel enabling privilege escalation
Zero-days across every major OS and every major browser

Five million automated hits. Zero catches. Mythos found it anyway.

This is the moment where deterministic security tooling hits its ceiling. Fuzzers, static analyzers, and symbolic execution work within the space of known patterns. Mythos reasons about code the way a senior security researcher does — contextually, creatively, following the logic of what could go wrong rather than what has gone wrong before.

The coalition: voluntary, industry-led, fast

Here's what I find most interesting about the structure of Glasswing: it's not a government program.

Amazon, Apple, Google, Microsoft, Nvidia, Cisco, CrowdStrike, JPMorgan, the Linux Foundation — 12 launch partners and 40+ additional organizations. No mandate. No regulatory requirement. No bureaucratic committee that spent 18 months drafting a framework.

Just a coalition of companies that looked at the threat landscape and decided that waiting for regulators to catch up was a losing strategy.

That's how it should work. Industry moving faster than policy, with enough transparency to be accountable. Within 90 days, Anthropic will publish a full report: vulnerabilities found, patches shipped, security improvements achieved. Public. Open. Reproducible.

The $100 million in model credits and $4 million in direct donations to open-source security organizations (OpenSSF, Apache Software Foundation) are real money going to the infrastructure that holds the internet together. The Linux kernel, FFmpeg, OpenBSD — these aren't niche tools. They run banks, hospitals, power grids.

Why this matters to every engineer reading this

I've spent years building software. Automated tests, CI/CD pipelines, security scans baked into the deploy process. I thought that was enough.

It's not enough anymore.

The attack surface of modern software is too large, too interconnected, and too old. We have code that's been running in production since before most of our current team members learned to code. Nobody fully understands it. Nobody has audited all of it. And now an attacker with a capable AI model can map that entire surface faster than your team can read the output.

Project Glasswing doesn't solve this alone. But it's the first time the defenders have access to the same class of capability as the attackers.

That matters. For years the asymmetry ran the other way — attackers needed to find one hole, defenders needed to protect everything. AI didn't change that asymmetry. It accelerated it. Until now.

What Glasswing establishes is a new precedent: frontier AI applied defensively, at scale, to real production software, with public accountability. If it works — and the early results suggest it does — it becomes the new baseline expectation for how serious organizations manage security.

In two years, "did you run this through an AI vulnerability analysis?" will be as standard as "did you write tests?"

The techno-optimist case

I'm not naive about dual-use. The same model that found that 27-year-old OpenBSD bug could, in the wrong hands, have weaponized it. Anthropic knows this. That's exactly why Mythos Preview is restricted.

But I'm also not willing to accept the pessimist framing that says AI in security is inherently destabilizing. That framing ignores something important: the bad actors don't wait for permission. They're already using every available model to find vulnerabilities. The question isn't whether AI changes the security equation — it already has. The question is whether the people building critical software will have access to equivalent tools.

Glasswing says: yes, they will.

That's the techno-optimist case. Not that technology solves everything automatically. But that when the people who understand the stakes make deliberate, coordinated decisions, technology can tip the balance toward the people trying to protect rather than the people trying to destroy.

The good guys finally have an AI that's faster than the attackers.

That's worth paying attention to.

Follow me on X: @crisesarmiento

Sources:

Love seeing a full FOSS stack (Godot/Blender/MuseScore/Audacity/Kdenlive) powering a solo project, plus the candid lessons about pipelines + UI scaling. Great reminder to treat the first game as education.

Cristian Sarmiento — Tue, 24 Feb 2026 16:15:07 +0000

Tom Nijhof

Jan 4

A Solo Game Developer who uses only Free & Open Source Tools

#programming #gamedev #godot #blender

Comments 3

5 min read

Analysis of Windsurf’s Target Audience Shift

Cristian Sarmiento — Thu, 15 May 2025 20:42:28 +0000

Introduction
Based on https://windsurf.com/blog/windsurf-wave-9-swe-1?_sc=ODEwMDk4MiM0OTY1OTMz

This document analyzes the shift in Windsurf’s target audience, as initially proposed by the user: “Cursor is pointing to studes and vibe coders, while Windsurf is becoming a tool more aiming to software engineers that think more like a system engineer or architect.” The analysis evaluates whether Windsurf, an AI-powered integrated development environment (IDE), is moving away from students and less experienced developers toward professional software engineers, particularly those with system engineering or architectural mindsets, and explores the fundamentals supporting this shift.

Initial Analysis
The user suggests that Windsurf, formerly Codeium, is repositioning itself to cater to software engineers who approach development with a system engineer or architect’s perspective, focusing on large-scale, complex systems. In contrast, Cursor, another AI-powered IDE, targets “studes and vibe coders,” interpreted as students and less experienced developers or hobbyists focused on coding culture. This implies Windsurf is prioritizing enterprise-grade features and professional use cases over accessibility for beginners.

Fundamentals Supporting the Analysis

1. Windsurf’s Enterprise Focus
Windsurf has significantly expanded its enterprise customer base, serving over 1,000 enterprises, including Zillow, Dell, JP Morgan Chase, and Broadcom. According to a Contrary Research report (April 30, 2025), Windsurf’s enterprise product generates eight figures in annual recurring revenue (ARR), highlighting its focus on large organizations. Features like fine-tuned models, robust security, and analytics for ROI transparency cater to system engineers and architects managing complex, secure systems in enterprise settings.

2. Introduction of Frontier Models
The blog post “SWE-1: Our First Frontier Models” (May 15, 2025) introduces Windsurf’s SWE-1 model family, designed for the entire software engineering stack, not just coding. These models offer deep contextual awareness and multi-file editing, enabling coherent changes across large codebases. Such capabilities align with the needs of system engineers and architects who work on interconnected, production-level systems, as noted in a DeepLearning.AI course (February 25, 2025).

3. Productivity Metrics for Professional Developers
Windsurf emphasizes productivity metrics relevant to enterprise environments, such as a 40-200% increase in developer productivity, 4-9x decrease in onboarding time, and reduced PR cycle times. These metrics, highlighted on Windsurf’s website, are critical for professional engineers managing large-scale projects, where efficiency and standardization are paramount. A CodeParrot.ai guide (February 2, 2025) underscores Windsurf’s ability to maintain development flow, a priority for experienced developers.

4. Shift from Free Tier Accessibility
Historically, Windsurf (as Codeium) offered a free tier with broad IDE compatibility, appealing to students and hobbyists, as noted in a DEV Community post (November 15, 2024). However, recent developments, including the Pro Ultimate plan ($60/month) for large teams and FedRAMP High authorization (March 2025), indicate a pivot toward enterprise-grade solutions. This shift reduces accessibility for casual users while enhancing Windsurf’s appeal to professional developers.

5. Comparison with Cursor
Cursor, a Visual Studio fork like Windsurf, targets a broader, less technical audience, including beginners and hobbyists, as discussed in a Reddit thread (November 21, 2024). Cursor’s user-friendly interface supports rapid prototyping and individual projects, whereas Windsurf’s advanced features, such as Cascade’s contextual awareness and agentic capabilities, cater to complex, enterprise-level development. This contrast supports the user’s claim that Cursor targets “studes and vibe coders,” while Windsurf focuses on professional engineers.

6. Industry Recognition and Acquisition
Windsurf’s inclusion in Forbes AI 50 lists (2024 and 2025) and its $3 billion acquisition by OpenAI (May 6, 2025) underscore its leadership in AI-powered development tools. The acquisition, reported by Mashable, positions Windsurf as a key player in enterprise AI coding, aligning with the needs of system engineers and architects who require cutting-edge, secure tools.

Evidence and Insights

Enterprise Adoption: Windsurf’s client portfolio and enterprise-focused features (e.g., SOC2 Type 2 compliance, zero-data retention) indicate a strategic focus on large organizations, as detailed in Windsurf’s enterprise page (April 2, 2025).
Advanced AI Capabilities: ** The SWE-1 models’ ability to handle complex tasks, as described in a VentureBeat article (May 9, 2025), supports Windsurf’s suitability for professional developers working on large-scale systems.
**User Experience Differentiation: Windsurf’s focus on user experience (UX), as noted in the Contrary Research report, sets it apart from competitors like Salesforce Codegen, appealing to engineers who value seamless integration and contextual suggestions.
Market Positioning: Posts on X (May 15, 2025) highlight SWE-1’s optimization for software engineering, with performance nearing frontier models, reinforcing Windsurf’s professional orientation.

Counterpoints and Limitations

Free Tier Retention: Windsurf still offers a free tier, which could attract students and hobbyists, potentially diluting its enterprise focus.
Marketing Gaps: The Contrary Research report notes Windsurf’s marketing strategy lags in engaging software quality professionals, security experts, and product managers, which could limit its enterprise adoption.
Competition: Microsoft’s dominance in enterprise tools (e.g., GitHub Copilot, Visual Studio) poses a challenge, as large companies may prefer established vendors, as discussed in The Pragmatic Engineer (November 26, 2024).

Conclusion
Windsurf is shifting its target audience toward software engineers who think like system engineers or architects, moving away from students and less experienced developers. This shift is driven by its enterprise customer base, advanced frontier models, productivity metrics, and strategic focus on complex development tasks. While Cursor caters to a broader, less technical audience, Windsurf’s evolution positions it as a premier tool for professional, enterprise-grade development. Continued investment in enterprise features and marketing to diverse stakeholders will be critical for Windsurf to solidify this repositioning.

From Satellites to Software: Why Technical Interviews Need a Reboot

Cristian Sarmiento — Fri, 07 Mar 2025 10:45:55 +0000

Introduction

I’m Cristian Sarmiento, a Full Stack Engineer with over a decade of experience in software development, from building satellite simulators to designing scalable healthcare solutions. My career is a testament to solving real-world problems—complex, unpredictable challenges that demand hands-on ingenuity. I don’t have a degree; I learned my craft by doing, diving into projects and figuring things out as I went. This approach has fueled my success, but it’s also why technical interviews feel like an insurmountable hurdle. They’re built for people who excel at memorizing algorithms and reciting theory, not for someone like me who thrives on practical problem-solving.

As the tech industry evolves—especially with AI reshaping our roles—it’s time to rethink how we evaluate talent. In this article, I’ll share my journey, explore the science behind my learning style, and propose changes to the interview process that align with the future of tech and the human skills that matter most.

My Journey: A Decade of Hands-On Innovation

My career spans a diverse range of roles and industries, all unified by a passion for building solutions that work. At CareJourney by Arcadia, I led a team to develop serverless APIs and micro frontends using Nest.js and React, delivering secure, scalable healthcare solutions. I managed AWS infrastructure to ensure compliance and collaborated with cross-functional teams to integrate technologies seamlessly, improving efficiency and user experience.

Before that, at Darwoft, I created a QA automation area from scratch, rooted in an architectural principle of integration. What started as a solo effort grew into a 40-person team, with a client now relying on the architecture I designed. I also worked as a solo Python developer for a startup client, building Django applications, designing databases (Postgres, DynamoDB, Elasticsearch), and using JIRA APIs to train models that boosted team efficiency.

At INVAP, I spent years on satellite projects like SAOCOM 1A, 1B, ARSAT-1, and ARSAT-2. I developed C++ functionalities for the SAOCOM Mission Simulator, created a Python + Qt GUI API, and managed flight operations for power and thermal subsystems. My work ensured these satellites performed optimally in orbit—real stakes, real results.

Earlier roles at CDSI, Nimbuzz, and Globant honed my skills in quality assurance and testing across platforms, from mobile games to messaging apps. Each project taught me something new, not through textbooks, but through trial, error, and iteration.

These experiences showcase my ability to tackle complex challenges and deliver tangible outcomes. Yet, despite this track record, technical interviews remain a stumbling block.

The Struggle with Technical Interviews

Technical interviews don’t reflect how I work. I’ve stood at whiteboards, asked to implement a binary search tree or explain an algorithm under pressure, and felt my mind lock up. It’s not that I don’t know these concepts—I’ve applied them in real projects for years. But I learned them by using them, not by memorizing them in isolation. My strength lies in building systems, debugging failures, and collaborating on solutions, not in performing theoretical exercises on demand.

Traditional interviews reward quick recall and textbook fluency—skills that don’t always translate to the job. For someone like me, who learns through experience, this setup highlights my weaknesses rather than my strengths. It’s frustrating to know I can design architecture for a 40-person team or keep a satellite in orbit, yet struggle to prove my worth in a 60-minute coding quiz.

The Science Behind Learning by Doing

My learning style isn’t a fluke—it’s backed by science. Constructivist learning theory, pioneered by Jean Piaget, argues that we construct knowledge through active engagement with our environment. It’s not about absorbing facts; it’s about building understanding through experience. That’s how I’ve approached every project—diving in, testing ideas, and refining solutions.
David Kolb’s experiential learning model takes this further, positing that knowledge emerges from transforming experience into practical insights. Research supports this: studies on project-based learning in STEM fields show that hands-on engagement leads to better retention and application of knowledge compared to traditional lectures. My 10 years of building systems—from satellites to software—mirror this process. I’ve learned more from debugging a failing API than I ever could from a classroom.

AI and the Shifting Industry Landscape

The tech world I entered a decade ago is unrecognizable today, thanks to AI. Routine tasks are being automated, and the focus is shifting to skills AI can’t replicate: critical thinking, creativity, and adaptability. The World Economic Forum’s Future of Jobs Report (2020) underscores this, listing these abilities as top priorities for the future workforce.

This shift aligns perfectly with my strengths. My career is built on adapting to new challenges—whether it’s integrating AI tools at Darwoft or optimizing satellite operations at INVAP. As AI handles the grunt work, engineers need to experiment, learn quickly, and pivot as technologies evolve. Memorizing algorithms matters less when you’re working alongside systems that adapt in real time. My hands-on experience feels more relevant than ever, yet the interview process hasn’t caught up.

Reimagining the Interview Process

To find talent suited for this AI-driven future, we need to overhaul technical interviews. Here’s how, based on my experience and the industry’s trajectory:

1. Replace Whiteboards with Real Projects
Ditch the on-the-spot coding tests. Give me a practical challenge—like building a small application or fixing a bug—and a few days to solve it. This mirrors real work and lets me showcase my ability to deliver results.
2. Focus on Problem-Solving, Not Memorization
Ask how I’d tackle a problem, not whether I can recite an algorithm. Let me walk through my process—how I analyze, experiment, and iterate. In an AI world, how I think matters more than what I’ve memorized.
3. Incorporate Pair Programming
Most of my work involves collaboration. A pair programming session would reveal how I code, communicate, and adapt in real time—without the artificial pressure of a solo test.
4. Evaluate Past Work
My projects—like the QA automation architecture at Darwoft or the SAOCOM simulator—say more than any interview could. A portfolio review or deep dive into my contributions would highlight my practical impact.
5. Test AI Adaptability
Ask how I’ve used AI tools or learned new technologies. Assess my ability to experiment and grow with the systems shaping our field. That’s a skill you can’t measure on a whiteboard.
These changes would better evaluate engineers like me, who excel in practice but falter in theory-heavy settings. They’d also uncover talent ready to thrive in a world where AI is a partner, not a replacement.

Conclusion

I’ve spent over 10 years proving that hands-on experience can outshine a degree. From satellites to healthcare software, I’ve built a career on solving real problems, even if technical interviews don’t always recognize it. Science validates my approach: learning by doing is powerful. And with AI transforming tech, the skills I’ve honed—adaptability, problem-solving, persistence—are more vital than ever.
It’s time for interviews to evolve. By prioritizing practical skills over theoretical drills, companies can tap into a broader talent pool and build teams equipped for the future. I’m not asking for a shortcut—just a chance to prove myself the way I always have: by getting my hands dirty and making things work.

References

Piaget, J. (1950). The Psychology of Intelligence.
Kolb, D. A. (1984). Experiential Learning: Experience as the Source of Learning and Development.
Krajcik, J. S., & Blumenfeld, P. C. (2006). Project-based learning.
World Economic Forum (2020). The Future of Jobs Report.

From Vision to Reality: How TestSenseAI's GitHub Actions Library Revolutionises Test Framework Development

Cristian Sarmiento — Sun, 12 Jan 2025 11:11:29 +0000

As a founder building TestSenseAI in 2025, I've become increasingly convinced that the future of test automation isn't just about writing better tests – it's about fundamentally reimagining the entire development lifecycle. Today, I want to share a breakthrough that's transforming how we approach test framework development: our innovative GitHub Actions library for requirements management and development criteria.

The Missing Link in Modern Testing

When I first started building TestSenseAI, I noticed a critical gap in the testing ecosystem. While we had sophisticated tools for writing and executing tests, the bridge between requirements and implementation remained surprisingly manual and error-prone. This realisation led to one of our most powerful innovations.

The Power of Automated Requirements Engineering

Our GitHub Actions library isn't just another tool – it's a paradigm shift in how we think about framework development. Here's what makes it revolutionary:

1. Intelligent Requirements Processing

We've built a system that transforms plain markdown requirements into structured, actionable framework components. This isn't just about parsing text – it's about understanding intent and context:

Smart Feature Detection: Our system automatically identifies key framework components from natural language descriptions
Context-Aware Analysis: Requirements are analyzed in the context of existing framework capabilities
Automated Dependency Mapping: Dependencies between components are automatically identified and tracked

2. Revolutionary Development Workflow

What truly sets us apart is how we've integrated this into the development workflow:

This workflow automation means:

Zero Friction: Requirements flow seamlessly into development tasks
Perfect Traceability: Every feature maps back to its original requirement
Continuous Validation: Requirements are continuously validated against implementation

Strategic Advantages in Framework Development

The impact of this approach extends far beyond just efficiency:

1. Accelerated Innovation Cycle

By automating the requirements-to-implementation pipeline, we've drastically reduced the time from concept to working code. This means:

Faster iteration cycles
More responsive framework development
Rapid adaptation to emerging testing needs

2. Enhanced Quality Assurance

Our approach inherently builds quality into the framework development process:

3. Future-Proof Architecture

Perhaps most importantly, this system is designed to evolve:

Why This Matters for 2025 and Beyond

In 2025's landscape of accelerating development cycles and increasing complexity, this approach isn't just nice to have – it's essential. Here's why:

Development Velocity: In a world where software changes by the hour, manual requirements management is a bottleneck we can't afford.
Quality at Scale: As systems grow more complex, maintaining quality through manual processes becomes impossible.
AI Integration: Our architecture is built from the ground up to leverage and evolve with AI advancements.

Looking Ahead: The Framework Development Revolution

This isn't just about building a better testing framework – it's about revolutionizing how frameworks are developed. Our GitHub Actions library represents a fundamental shift in how we think about the relationship between requirements and implementation.

Conclusion: Leading the Future of Framework Development

As we look toward the future, it's clear that the way we build frameworks must evolve. At TestSenseAI, we're not just preparing for this future – we're actively creating it. Our GitHub Actions library isn't just a tool; it's a glimpse into the future of software development.

For teams building testing frameworks in 2025 and beyond, the question isn't whether to automate requirements and framework development – it's how quickly they can adopt this new paradigm. At TestSenseAI, we're proud to be leading this revolution, creating tools that don't just solve today's problems but anticipate tomorrow's challenges.

The future of framework development is intelligent, automated, and seamlessly integrated. And with TestSenseAI's GitHub Actions library, that future is already here.

About the Author: As the founder of TestSenseAI, I'm passionate about revolutionising how we think about test automation. With over a decade of experience in software testing and framework development, I'm now focused on building the next generation of intelligent testing tools.

New Year, New Era: Why 2025 Won’t Be Business as Usual [by an autistic Software Architect]

Cristian Sarmiento — Thu, 02 Jan 2025 11:57:21 +0000

Happy New Year! Another 365 days (366 for 2024, yeap, it felt longer) have passed—and if you’re reading this, you’ve probably already scrolled past dozens of “New Year, New Me” posts. But here’s the thing: 2025 isn’t just another year. The world of work, tech, and talent is changing faster than we can say “meritocracy.” And not everyone is going to make it across the finish line unscathed.

The Myth of “Fun” Workplaces

Let’s face it: many companies pride themselves on “fun” HR gimmicks. They clutter the office with loud games, host birthday parties every other day, and hand out pointless certificates—while ignoring the real reasons talented people come to work: to grow, innovate, and make a difference.

I’m not against a bit of fun (we’re humans, not robots), but in some organizations, these activities overshadow the very essence of productivity. You know the scene: a superficial “culture of inclusion,” complete with rainbow logos in June, and zero concrete effort to elevate diverse viewpoints the rest of the year.

The Cost of Mediocrity

We all know at least one bright, hardworking person who’s underutilized—or worse, discriminated against—because their honesty or directness is perceived as too abrasive. Often, these individuals are multilingual, highly skilled, and passionate about exceeding expectations. Meanwhile, the folks who’ve only dabbled in JavaScript or got an AWS certificate they’ve never used somehow climb the corporate ladder with ease.

Why does this happen? Because mediocrity thrives where superficial engagement is rewarded. Instead of championing innovation, some workplaces champion the illusion of teamwork: team-building games that yield zero practical outcomes, endless standups that substitute for real communication, and project managers who master the art of scheduling but rarely solve actual problems.

Why 2025 Is Different

You’ve likely noticed the buzz around artificial intelligence. Everyone and their uncle is talking about how it will “change everything,” and for once, the hype might be justified. Here’s the catch: AI won’t just automate mundane tasks—it might also expose underqualified roles that add little value. Think of it as a massive performance review for the entire job market.

Rote tasks will be replaced by algorithms.
Shallow knowledge will get overshadowed by those who’ve actually mastered their craft.
“Busy-work specialists” may soon realize that software can do their checklists faster—and without complaining.

The real winners in 2025? The people who actually use their skills every day, who adapt quickly, and who think critically. In other words, the very folks who often feel stifled by the current, superficial culture—especially those in neurodivergent communities who bring unique perspectives, high focus, and a drive for excellence.

Turning Frustration into Hope

If all this sounds harsh, here’s the good news: this shake-up can be a catalyst for positive change. I truly believe in a future where results, dedication, and authenticity speak louder than office politics. That means:

Organizations willing to adapt will finally empower real talent.
Neurodivergent employees, bilingual specialists, and critical thinkers will find roles that value them for what they do best.
Innovation will flourish when we end the era of “look busy, get promoted.”

Yes, it’s easy to feel frustrated. But as we ring in the new year, let’s remember that with disruption comes opportunity—especially for those who’ve felt overlooked. This is particularly true for autistic individuals who throw themselves wholeheartedly into work, not just to fulfill a role but to excel in it. For many of us, work is more than a paycheck; it’s a channel for our passions, our dedication, and our genuine desire to create something meaningful.

Let’s greet 2025 with the understanding that not everyone will (or should) survive in the old way of doing business. Companies that continue to ignore the true value of their best people—or that mistake shiny HR activities for genuine inclusion—will learn the hard way.

The Year of Authentic Growth

So here’s my New Year’s toast:

May we finally see workplaces where skill outshines superficial labels.
May artificial intelligence serve as a tool that amplifies human creativity, not a threat to genuine talent.
May every talented individual who’s been overlooked or misunderstood find a place to shine.

2025 is knocking, and I, for one, am excited. Because for those of us who’ve been speaking the truth all along—and putting in the real work—this might just be the year that everyone else catches up.

Happy New Year. Let’s make sure it’s not like the others.