Developer news without the noise — My Dev News in Telegram.
Tag
All blog posts with this tag.
A practical breakdown of token costs in AI-assisted programming: input and output, context, caching, agent loops, Cursor versus API, and how to measure the price of working code, not just a per-million-token rate.
AI agents are changing software development: an experienced programmer can assign implementation, tests, review, and docs in parallel. A working model, hard limits, and the skills of an engineer who directs workers — not only types code.
Microsoft, GitHub, and METR: AI can cut task time by tens of percent—or slow mature repos by ~19%. Generation speed, quality, and the productivity illusion.
Agent PRs outpace review capacity: review debt, what AI can catch vs what humans must own, a post-diff checklist, human gates, and a team playbook for light vs deep review in 2026.
A practical model for agentic development: context, task decomposition, verification loops, multi-agent orchestration, security, DORA metrics, and the engineer's role in the 2026 SDLC.
