/vcg/ — Vibe-coding general
Active thread on vibe coding, agentic engineering, AI IDEs and shipping with LLMs; recent posts discuss tools and limitations.
Monday, July 13, 2026
/vcg/ — Vibe-coding general — Monday, July 13, 2026
Terence Tao shares recent experiments porting old Java applets and building new visualizations using AI agents and vibe coding; DevOps.com explores integration of vibe coding into CI/CD workflows; ongoing 4chan /vcg/ thread and X posts highlight new tools and artist applications. Vibe coding continues to gain traction as an AI‑assisted, flow‑oriented development style, but concerns about code quality, security, and long‑term maintainability are becoming more prominent in the discourse.
Biggest developments
Old and new apps, via modern coding agents
Ported two dozen old Java applets to modern JS with AI assistance in hours; created new spacetime diagram and Gilbreath conjecture visualizers via vibe coding sessions.
Vibing with the Future: Why “Vibe Coding” Is the Next Big Wave for DevOps and CI/CD Teams
Vibe coding aligns AI with team culture, standards and preferences to streamline CI/CD, reduce onboarding and improve code consistency.
Important posts & threads
vibe coding is opening a whole new lane for artists
vibe coding is opening a whole new lane for artists, curious to see what weird stuff comes out of it
Vibe Codingから Software 3.0へ
Zenn article details moving from one-off instructions to team-wide rules in CLAUDE.md and sub-agents for consistent vibe coding output.
Built a product to help freelancers speedrun vibe coding jobs
App searches Instagram/Facebook/Google maps for businesses without websites and auto-creates sites from leads.
Bullish takes
- Artists gaining new creative lanes via vibe coding
- DevOps teams can achieve deeper alignment and faster CI/CD with vibe-aware AI
- Vibe coding is accelerating prototyping and feature delivery by offloading implementation to AI agents, enabling developers to ship faster while staying in flow state.
- New AI‑native platforms and vertical LLMs (e.g., Base44’s Base1) are tailoring vibe‑coding workflows to specific domains, lowering the barrier to entry for non‑expert builders.
- Organizations are beginning to formalize code‑quality guardrails for vibe coding, suggesting the practice is maturing beyond ad‑hoc experimentation into a structured engineering pattern.
Critical takes
- Vibe coding is associated with growing technical debt and review debt, as rapid AI‑generated code can bypass traditional design and security reviews.
- Community discussions highlight recurring issues such as missing authorization checks, exposed API keys, and insecure debug endpoints, raising concerns about long‑term maintainability and security.
- Some practitioners describe vibe coding as a ‘weird phase’ of software development, implying that current practices may be unstable or unsustainable without stronger process discipline.
Why this matters
Fresh examples show vibe coding scaling from personal experiments to team workflows and specialized tools.