‹ Series
Claude code — Complete Coverage
All articles about claude code on TamilTech
Articles in this series
-
1
Anthropic Blocks Claude Code Access in Third-Party Tools Like OpenCode: Developers React
Anthropic has implemented strict technical safeguards preventing third-party applications like OpenCode from accessing Claude AI models through Claude Code credentials, sparking controversy among developers paying $200/month for Max subscriptions.
-
2
OpenAI Codex Just Got Plugins — GitHub, Gmail, Vercel in One Click. Is It Enough to Catch Claude Code?
OpenAI just added a plugin system to Codex — its agentic coding assistant — letting developers connect GitHub, Gmail, Box, Cloudflare, Vercel and more with one-click installs. It's a direct move to close the gap with Claude Code, which already has this. But will developers actually switch? Here's the real picture.
-
3
Warning for Indian Developers: Searching for Claude Code's Leaked Source? Those GitHub Repos May Have Malware Inside
Anthropic accidentally leaked Claude Code's entire source code — 513,000 lines of TypeScript — on March 31. Hackers immediately flooded GitHub with fake repositories pretending to host the leaked code, but actually delivering Vidar infostealer and Ghostsocks malware. If you're a developer who went looking, read this now.
-
4
HumanX Recap: Claude Code steals the show, China’s open‑weight surge catches eyes
HumanX revealed that Anthropic’s Claude Code dominated conversations, while several execs warned that China is sprinting ahead with open‑weight AI models.
-
5
AI Coding War 2026: Claude Code vs GPT-5.2 Codex vs Gemini Code Assist vs DeepSeek Coder
The AI coding landscape has exploded in early 2026 with major updates from every provider. Claude Code 2.1.0, GPT-5.2-Codex, Gemini Code Assist, and DeepSeek V4 are all vying to become the developer's go-to AI assistant.
-
6
Anthropic Accidentally Leaked Claude Code's Source — and It Reveals a 24/7 AI Agent Called Kairos
Anthropic accidentally published 500,000+ lines of Claude Code's internal source code to a public registry. Developers digging through it found hidden features including 'Kairos' — a persistent AI that runs in the background even when you close the app — and 'AutoDream,' which lets the AI consolidate memories while you sleep. Here's what leaked and what it means.
-
7
Anthropic Just Blocked Claude Code Users From Using OpenClaw for Free — And the Timing Is Suspicious
If you're a Claude Code subscriber using OpenClaw or any other third-party AI coding harness, your subscription no longer covers it. From April 4, 2026, Anthropic requires separate pay-as-you-go payment for third-party tool usage. The reason given is compute costs. The timing — right as OpenClaw's creator just joined OpenAI — is raising a lot of eyebrows.
-
8
Claude Code Vanishes from Anthropic Pro? What’s the Real Story
Users noticed Claude Code disappearing from Anthropic’s Pro plan. The company says it’s a small rollout test. Here’s what’s really happening and what it means for Indian devs.
-
9
Sarvam Launches Arya: The Production-Ready AI Agent Stack Built in India — Immutable Ledger, Claude Code Integration, and Why Frameworks Are Not Enough
Sarvam AI launched Arya on February 10, 2026 — an agent orchestration stack designed as infrastructure, not a framework. With 8 flat primitives, an immutable append-only Ledger, Agrippa task graph exposed as MCP, and Claude Code integration via Acta MCP (spawning 4 parallel debugging subagents), Arya offers composable guarantees for production AI agents.
-
10
Warning: Hackers Are Using the Claude Code Leak to Infect Developers' Computers With Malware
Anthropic accidentally leaked Claude Code's entire source code on March 31. Hackers immediately set up fake GitHub repositories pretending to host the leak — and anyone who downloaded from them got hit with Vidar infostealer malware that steals passwords, credit card data, and browser history. If you searched for 'leaked Claude Code' on Google this week, read this now.
-
11
AMD's AI Director Just Proved Claude Code Is Getting Dumber — 6,852 Sessions of Data Don't Lie
AMD's Stella Laurenzo analyzed 6,852 Claude Code sessions and 234,760 tool calls. The verdict: Claude Code's performance has degraded significantly since March 2026. Stop-hook violations jumped from zero to 10 per day, reading behavior dropped from 6.6x to 2x per file, and the AI started doing full rewrites instead of targeted edits. Here's what went wrong and what it means for developers who depend on AI coding tools.
-
12
Claude Code Security Woes: Are Indian Developers at Risk of Secret Leaks?
Indian developers are raising concerns about Claude Code potentially leaking sensitive code and secrets, sparking debates about AI tool security in the Indian tech ecosystem.
-
13
Claude Code Flaws Allow Remote Code Execution and API Key Exfiltration — Check Point Research Exposes Critical Vulnerabilities
Check Point Research has discovered critical vulnerabilities in Anthropic's Claude Code: CVE-2025-59536 (Hooks and MCP exploit allowing remote code execution) and CVE-2026-21852 (API key exfiltration via ANTHROPIC_BASE_URL override). A single malicious commit in a repository can compromise any developer who clones and opens it in Claude Code. Hooks execute automatically without consent, MCP servers can bypass safeguards, and API keys are exposed in plaintext.
-
14
OpenAI Just Launched a $100/Month ChatGPT Plan — Here's Why It's a Direct Shot at Claude Code
OpenAI launched a brand new $100/month ChatGPT plan on April 9, 2026 — sitting right between the $20 Plus tier and the $200 Pro Max tier. The pitch: 5x more Codex usage, unlimited Instant and Thinking models, and exclusive access to new models. This is OpenAI directly punching at Anthropic's Claude Code, which also costs $100/month. Sam Altman says Codex has hit 3 million weekly users.
-
15
AWS Launches Agent Plugins: Deploy to AWS in 10 Minutes with Claude Code and Cursor
AWS has open-sourced Agent Plugins — a framework that equips AI coding agents like Claude Code and Cursor with skills to architect, deploy, and operate on AWS. The deploy-on-aws plugin transforms "deploy to AWS" into a 5-step automated workflow generating CDK/CloudFormation with cost estimates.
-
16
Claude Code Leak Through npm: What It Means for AI Developers
Anthropic’s Claude source code surfaced on an npm registry, sparking security concerns and a scramble among AI developers. Here’s the full story, the Indian angle, and what you should do next.
This series is still growing. New parts land here first.
