Selected writing
Writing that feeds the handbook.
Habr
Short prompt does not mean cheap prompt
Agent loops, tool list stability, allowed tools and cache-aware prompt design.
Habr
7 prefix cache anti-patterns
Timestamp drift, floating tool order, round-robin routing and KV-cache lifetime.
Habr
Effective cost with cache
Why model choice needs cache-aware economics, not only list prices.
Habr
Agent Skills are more than a prompt folder
How agent skills relate to tools, RAG, MCP and agent architecture.
Habr
Why one million tokens do not solve context
Working context, quality decay and the limits of long context windows.
Habr
AI agents explained with Deep Research
How LLM calls, workflows, RAG and agent loops differ in practice.
Habr
How answer style changes LLM rankings
Why model choice should not be reduced to polished answer style.
Habr
How to choose an LLM for a product
Context, quality, cost, jurisdiction and other model-selection parameters.
Telegram
AI da paren!
Notes on AI platform engineering, tools, vLLM, agent systems and production trade-offs.
Habr articles
Long-form technical writing on cache, cost, inference and production AI scenarios.
Telegram channel
Short notes, links, observations and draft thinking about production AI platforms.