Kedasha K.
Chicago, Illinois, United States
9K followers
500+ connections
View mutual connections with Kedasha
or
New to LinkedIn? Join now
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
View mutual connections with Kedasha
or
New to LinkedIn? Join now
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
View Kedasha’s full profile
-
See who you know in common
-
Get introduced
-
Contact Kedasha directly
Other similar profiles
Explore more posts
-
Howdang Rashid
Powercademy • 31K followers
The Copilot Studio CAT team just open-sourced something that genuinely excites me. It's called "Skills for Copilot Studio" - a plugin for Claude Code and GitHub Copilot CLI that lets you author, test, and troubleshoot Copilot Studio agents as YAML. From your terminal. With natural language. Three commands: /𝗰𝗼𝗽𝗶𝗹𝗼𝘁-𝘀𝘁𝘂𝗱𝗶𝗼:𝗮𝘂𝘁𝗵𝗼𝗿 - builds topics, actions, knowledge sources . /𝗰𝗼𝗽𝗶𝗹𝗼𝘁-𝘀𝘁𝘂𝗱𝗶𝗼:𝘁𝗲𝘀𝘁 - runs test utterances against published agents . /𝗰𝗼𝗽𝗶𝗹𝗼𝘁-𝘀𝘁𝘂𝗱𝗶𝗼:𝘁𝗿𝗼𝘂𝗯𝗹𝗲𝘀𝗵𝗼𝗼𝘁 - debugs routing and disambiguation issues. Instead of clicking through the portal, you describe what you want and get (pretty good) YAML with the correct schema. The CAT team reports up to 20x faster development. I want to say a genuine well done to the CAT team for this. Not just the plugin, but what it represents. They're encoding their own internal best practices directly into tooling that everyone can use. It gives me a lot of confidence in the direction this technology is heading. I made a cheat sheet breaking down the full workflow, commands, and gotchas. It's in the Powercademy Success Kit that you can grab here: https://lnkd.in/gsKBUaM7 Still in beta, so review your YAML before pushing. But the foundation is solid. #CopilotStudio #PowerPlatform #ClaudeCode #MicrosoftAI #AgentDevelopment
317
11 Comments -
Billy Robins
Jellyfish • 10K followers
The new DORA Report is here from Google with research partners spanning GitLab GitHub IT Revolution SkillBench Workhelix. Jellyfish is ecstatic to be a partner. Nathen Harvey sharing some highlights at #ETLS2025 with his patented energy + humor. "Take out your phones now. I don't care what you do with your phones but it makes me feel better." Excited to dig into the different team profiles and the 7 key practices that drive better outcomes from adopting AI. The highlights about how AI amplifies your team's software development practices (for better or worse!) ring so true. Full report here 👉 https://bit.ly/429eLXi
68
1 Comment -
William Collins
Itential • 5K followers
🌶️ Another ride down that spicy hot take highway. I feel like this is actually a valid critique of spec-dumping MCP servers dressed up as a protocol indictment e.g., REST is bad because someone built an awful API. Just to add an alternate perspective - The points on token bloat are 100% valid (that's a design failure worth calling out). The leap from "some" MCP server is over-engineered to "MCP is enterprise middleware theatre" skips a few steps. As someone who literally LIVES in the CLI, I can attest to the CLI being great when the model has training signal and you control the environment. This falls apart rather quickly with internal systems, proprietary things, and structured error handling at scale. The answer to a bloated spec isn't parsing stdout, it's better MCP design. And let's also not forget that CLIs output unstructured data which is historically fragile. 💡 CLIs are great for hackers and humans; MCP is the contract we need when we start offloading mission-critical, multi-step workflows across proprietary systems without 'hallucinating' the command flags.
14
2 Comments -
Viktor Farcic
Upbound • 23K followers
The question isn't whether AI replaces creative work. It's which side you're on. Orchestrators sit next to their files, run proper agents, and plug in specialized tools (like Higgsfield). Everyone else gets replaced by them.👇 https://lnkd.in/e9UWf_fW #AIVideoGeneration #MCPIntegration #ClaudeCode
16
1 Comment -
Kat Cosgrove
Kubernetes • 4K followers
Maintaining an open source project is hard. Most maintainers are managing globally distributed contributors who are largely working for free, while companies build products and profit on top of their work. At the same time, those same companies expect bug fixes, features, and security updates on timelines that work for them, not for us. When something goes wrong, the blame usually lands on the maintainers. If your business relies on open source, helping maintain it isn’t optional. It isn’t charity. It’s part of a secure software supply chain. The health of these projects doesn’t just affect maintainers, it affects every company and product that depends on them. I spend a lot of my time at work contributing upstream as a Kubernetes maintainer and supporting the broader ecosystem. I’m very lucky to work at Minimus, a company that recognizes the importance of that work and considers it a large part of my job as a Developer Advocate. It’s so rare to see that in a smaller company. I wrote more about this here: https://lnkd.in/geZBAVXV
215
7 Comments -
Klint Finley
Dynatrace • 681 followers
Observability isn’t just about collecting data. It’s about connecting the dots across your stack to answer why something is broken, not just what is broken. For developers, this means the ability to ask ad-hoc questions about live systems and get answers fast enough to change code with confidence. Modern observability tools can help you catch issues before they become incidents, validate deployments, and understand how your code behaves in the real world. This excerpt from our new Developer’s Guide to Observability, walks you through how to build an observability workflow that works for you, not against you. https://lnkd.in/gG8YGrYa
1
-
Ed Summers
GitHub • 6K followers
The highlight of GitHub Universe for me was meeting Christian Grobmeier. Christian is a maintainer of Log4j, one of the many open source projects that power the technologies that humanity uses to work, learn, live, and play. Christian recently sat down with Gregg Cochran and shared an amazing behind-the-scenes perspective on Log4Shell, which was a critical security incident within Log4j. The interview is a must read and watch for everyone that builds technology. The Log4Shell incident was a watershed moment. It clearly identified the need to support open source maintainers. GitHub's Secure Open Source Fund fills this gap by providing security training and resources. Here's my favorite quote from the interview with Christian, “With this training, developers are no longer the weakest link. Instead, they’re the first line of defense.” Thank you, Christian, for your contribution and the courage to transparently share your perspective on Log4Shell. Your words have inspired me to think more deeply about how GitHub can help improve the accessibility of open source for ~1.3 billion people with disabilities globally. https://lnkd.in/evzRCB5h #GitHubUniverse #secuirity #accessibility
95
-
Morten Rand-Hendriksen
LinkedIn • 92K followers
My socials overflowed with developer fury this weekend over Anthropic's crackdown on 3rd party use of Claude subscriptions. Tl;Dr: Claude Max subscribers who log in via OpenCode CLI have been locked out as Anthropic consolidates use to Claude Code only. To understand why this is complicated, a real-world analogy is helpful: Is Anthropic limiting Claude subscribers to using Claude Code more like Amazon limiting Audible subscribers to only using the official Audible player to listen to books they bought, or like Netflix limiting subscribers to only using official Netflix apps to watch content? Here's a high-level breakdown of what happened and the response: - You can pay for your Claude model use in two ways: Using Claude Code with a Pro or Max subscription, or using any other tool through the pay-as-you-go API - Comparing the pay-as-you-go Claude API to what you get from the most expensive Claude Max tier ($200/month), several developers claim API users would be charged over $1000/month for the same tokens and inference - Tools like OpenCode CLI have used good old fashioned User Agent spoofing to allow Claude Max subscribers to use their subs outside Claude Code - Late last week, OpenCode users got an error stating "This credential is only authorized for use with Claude Code and cannot be used for other API requests." - On Friday, Anthropic confirmed it has "tightened our safeguards against spoofing the Claude Code harness." - Developers see this as Anthropic enforcing vendor lock-in to Claude Code What do you think? Is Anthropic in the right here, or is this a land grab? Let's talk in the comments! -- Links: - Initial report of OpenCode error (GitHub issue): https://lnkd.in/gqVscayP - Hacker News thread: https://lnkd.in/g2NAA7yd - VentureBeat article covering the story with links to Anthropic statements: https://lnkd.in/gSaKWq_b
9
4 Comments
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content