Evgeny Tolmachev
Technical notes on AI and engineering.
- A Short History of LLM Agentic Research (2022–2026)
How LLMs went from single-turn Q&A to agentic coding harnesses: chain-of-thought, tool use, context windows, reasoning models, and Claude Code. Part 5 of a series on working with Claude Code in a large brownfield enterprise codebase.
- Claude Code's Agentic Harness: Tools, Context, and Permissions
The mechanics behind Claude Code: MCP versus CLI tools, context-window budgeting, project and settings scoping, and the permission system. Part 1 of a series on working with Claude Code in a large brownfield enterprise codebase.
- Teaching Claude Code Your Standards: Memory, Skills, and Subagents
How memory, skills, and subagents manage knowledge in Claude Code — with the actual file structure and rule examples. Part 2 of a series on working with Claude Code in a large brownfield enterprise codebase.
- Opinions on Agentic Coding
Opinions on trust, verification, and time cost formed from real-world experience. Part 3 of a series on working with Claude Code in a large brownfield enterprise codebase.
- Using Claude Code Across the SDLC
How Claude Code skills map to SDLC phases in an enterprise codebase: configuration, specification writing, Go implementation, testing, review, .NET, SQL, and deterministic tooling. Part 4 of a series on working with Claude Code in a large brownfield enterprise codebase.