|
|
Hi there, this is your daily ☕️ DevShot.
|
|
|
In today's DevShot:
|
|
🖥️ VS Code adds side-by-side AI chats 🔥 Mojo is now open source 🤖 TrueForge open-sources Claude agent rival 🛡️ Harness AI agents patch vulnerabilities 📌 Azure Cosmos DB opens data to AI agents 🛠️ Microsoft ships Aspire 13.5 Plus: 🎁 5 other news you might like, 🛠️ 5 tutorials, 🧰 6 tools, and 📚 5 papers.
|
|
Want to get the most out of ChatGPT?ChatGPT is a superpower if you know how to use it correctly. Discover how HubSpot's guide to AI can elevate both your productivity and creativity to get more things done. Learn to automate tasks, enhance decision-making, and foster innovation with the power of AI.
|
🖥️ VS Code adds side-by-side AI chats
LINK
|
- VS Code 1.134 now lets you arrange multiple chats in a session into horizontal or vertical grid groups, dragging conversations or subagent chats side by side to compare results and monitor agent work simultaneously.
- A new prompt timeline in the Agents window transcript gutter marks each prompt as a dot, shows lines added and removed when a prompt changes files, and links directly to those changes for review.
- Ctrl+F now searches across Chat view, chat editors, and the Agents window, including off-screen content, with case matching, whole-word, and regex options, while HTML files can be set to open in the integrated browser for preview by default.
|
🔥 Mojo is now open source
LINK
|
- Mojo, the systems language from Modular, has shipped its compiler and toolchain under an Apache 2 license, following its 1.0 release last week and fulfilling an open-source promise made back in May 2023.
- Rather than becoming a Python superset as first pitched, Mojo pivoted in August 2025 to being its own language with Python-inspired syntax, optimized to make GPU programming less painful even without full compatibility with existing code.
- Modular now leans on AI-assisted coding tools to migrate Python to Mojo, betting that maturing tooling and ecosystem support will smooth the transition instead of relying on drop-in interoperability with the current Python codebase.
|
🤖 TrueForge open-sources Claude agent rival
LINK
|
- TrueFoundry has released TrueForge, an open source agent harness pitched as an alternative to Anthropic's Claude Managed Agents, letting engineers build, deploy, debug, and govern production AI agents on any model or MCP server at roughly half the operating cost.
- TrueForge ships with support for OpenAI, Anthropic, and 20+ other models, plus 40+ built-in tools, sandboxed execution, human-approval workflows, large-context handling, generative UI, and Tavily-powered web search, routing every model call and MCP interaction through TrueFoundry's AI Gateway for budget and guardrail enforcement.
- Teams run it on their own infrastructure with their own keys, routing each task by cost, latency, or quality; on 14 level-one and level-two tasks from DevRev's Enterprise-Bench, TrueForge came in 50% cheaper at similar accuracy by using fewer tokens and non-Anthropic models.
|
🛡️ Harness AI agents patch vulnerabilities
LINK
|
- Harness launched a set of AI agents that scan for software vulnerabilities, judge which are exploitable, and draft validated fixes, opening a pull request against the flawed function for developers to approve before anything ships.
- The pipeline runs a deterministic static scanner, then an AI layer that trims false positives and catches logic flaws like missing authorization checks, while a Zero-Day Agent flags newly disclosed bugs and virtual patching blocks exploitation in production without code changes.
- Harness said frontier models surface roughly 10 times more vulnerabilities than conventional scanners, and the agents share one set of reachability data to keep teams off unexploitable findings, aiming to shrink discovery-to-fix time from weeks to hours.
|
📌 Azure Cosmos DB opens data to AI agents
LINK
|
- Microsoft is embedding agentic AI into Azure Cosmos DB, letting AI tools inspect database structure, generate queries against real schemas, and execute them with developer approval directly inside the workflow.
- GitHub Copilot in VS Code can now sample documents from a container to learn its schema before writing queries and dropping them into the Cosmos DB Query Editor, though sampling consumes request units and needs consent.
- The Cosmos DB Agent Kit ships 100+ recommendations on data modeling, partition keys, indexing, and vector search across Copilot, Claude Code, Codex, Cursor, and Gemini CLI, plus an optional local MCP server via Cosmos DB Shell.
|
🛠️ Microsoft ships Aspire 13.5
LINK
|
- Microsoft has shipped Aspire 13.5, refreshing the dashboard's visual design and homepage while expanding how AppHosts interact with developers and adding more flexibility to how applications get deployed to Kubernetes and Azure.
- The Interaction Service now supports JSON and YAML file imports via a file picker, progress dialogs with cancellation, and named arguments, while WithTerminal() adds an experimental interactive terminal for shells and REPLs inside the dashboard.
- TypeScript AppHosts are now generally available with feature parity to C#, and breaking changes include renaming ServiceProvider to Services, deprecating the GitHub Models integration, and removing the dashboard's AI Assistant chat.
|
|
2 Free AI Courses. No Credit Card Needed.5,000+ professionals use Skill Leap to get ahead with AI. Right now, two of their best courses are completely free - Claude 101 and the 14-Day AI Boot Camp. Claude 101 covers prompting frameworks, Artifacts, file analysis, and real-world workflows in 19 lessons. The Boot Camp covers ChatGPT, Gemini, Midjourney, and prompt engineering in 16 lessons. Downloadable workbooks. LinkedIn certificate. Zero cost, no credit card, no catch.
🔗 Other News & Articles
-
Two very Exciting memory management Optimizations going Into Linux 7.3
LINK
-
AWS deprecated this EKS auth method. 81% of clusters still run it.
LINK
-
Replit launches ‘Free Mode’ with OpenAI, letting users build AI apps without burning credits
LINK
-
AWS DevOps Agent traces pipeline failures to GitHub commits
LINK
-
Harper argues against the Multi-System stack and Releases 5.2
LINK
|
|
📄 Papers & Resources
|
> Cloud drift alerts can be ranked by risk so operators see fewer of them, cutting alert volume by ~73% while still flagging ~94% of security-relevant infrastructure changes, without needing machine learning.
LINK
|
|
> Software supply-chain checks reveal that even reproducible builds often can't be independently verified across four major package ecosystems, because missing source and build records leave outsiders unable to confirm published software matches its code.
LINK
|
|
> Document compiler picker is an open benchmark of 1,784 files that tells writers which engine to trust, showing Tectonic stays consistent across operating systems while classic LaTeX tools swing 12 to 20 percentage points.
LINK
|
|
> Factory control code generated by AI can now be automatically tested against a live industrial system before it's declared done, roughly doubling the rate at which programs actually behave correctly, from ~31 points to ~52.
LINK
|
|
> Mobile app repair testing checks whether AI coding agents' fixes actually work on real installed phone apps, revealing a huge ~69-point gap between the best and worst agents on Android tasks.
LINK
|
|
|
|
🧰 Latest Tools
| ElevenLabs: turns any script into studio-quality speech in seconds, with 5,000+ voices across 70+ languages. Free to start. GENERATE FOR FREE |
|
Sim: build and deploy AI agents through a visual or code workspace, with 1,000+ integrations and support for every major LLM provider.
LINK
|
|
Humalike x Hermes: gives AI agents turn-taking, timing, and memory APIs so they know when to speak, wait, or interrupt during conversations.
LINK
|
|
Coldtea.ai: automates QA and monitoring with AI agents that detect bugs post-deployment, flag issues, and generate actionable fix tasks automatically.
LINK
|
|
Kane CLI: automates browser tests from plain-English descriptions, driving real Chrome from your terminal and returning clear pass or fail results.
LINK
|
|
Openbase: helps developers choose npm and open-source packages by comparing popularity, maintenance activity, and reliability metrics alongside real user reviews.
LINK
|
|
FetchSandbox: reproduces API bugs deterministically across 60+ APIs by simulating webhooks and retries, generating shareable receipt URLs for PRs without burning quota.
LINK
|
|
|
|
💬 How did you find today's edition?
We read every reply — just reply to this email and let us know how we can improve!
|
|
|