January 2025
31 published posts filed under January 2025. Each article shows its actual publication date.
-
Cursor: Agent mode as a faster, more honest pair programmer
A field-tested take on agent mode as a faster, more honest pair programmer with Cursor: what it rewards, where it breaks, and how to keep the workflow honest.
-
Google Gemini: Multimodal debugging with a single screenshot
A field-tested take on multimodal debugging with a single screenshot with Google Gemini: what it rewards, where it breaks, and how to keep the workflow honest.
-
ChatGPT Codex: Sandbox surprises in the first month of Codex
A field-tested take on sandbox surprises in the first month of Codex with ChatGPT Codex: what it rewards, where it breaks, and how to keep the workflow honest.
-
Claude Code: Planning mode as the training wheels we kept
A field-tested take on planning mode as the training wheels we kept with Claude Code: what it rewards, where it breaks, and how to keep the workflow honest.
-
GitHub Copilot: When inline completion earned its place in the daily loop
A field-tested take on when inline completion earned its place in the daily loop with GitHub Copilot: what it rewards, where it breaks, and how to keep the workflow honest.
-
Local LLMs with Ollama: Picking a local model for code without benchmarks to lean on
A field-tested take on picking a local model for code without benchmarks to lean on with Local LLMs with Ollama: what it rewards, where it breaks, and how to keep the workflow honest.
-
Agentic coding: Budgets as the first feature any agent had to have
A field-tested take on budgets as the first feature any agent had to have with Agentic coding: what it rewards, where it breaks, and how to keep the workflow honest.
-
MCP servers: Exposing read-only resources before we dared expose actions
A field-tested take on exposing read-only resources before we dared expose actions with MCP servers: what it rewards, where it breaks, and how to keep the workflow honest.
-
AI coding workflows: Small reversible steps as a rhythm we borrowed from elsewhere
A field-tested take on small reversible steps as a rhythm we borrowed from elsewhere with AI coding workflows: what it rewards, where it breaks, and how to keep the workflow honest.
-
Prompt engineering for code: Learning that acceptance criteria beat cleverness
A field-tested take on learning that acceptance criteria beat cleverness with Prompt engineering for code: what it rewards, where it breaks, and how to keep the workflow honest.
-
Cline and Roo Code: Bring-your-own-key in the early days of model pricing
A field-tested take on bring-your-own-key in the early days of model pricing with Cline and Roo Code: what it rewards, where it breaks, and how to keep the workflow honest.
-
JetBrains AI Assistant: Junie arrives: the first autonomous task we trusted
A field-tested take on Junie arrives: the first autonomous task we trusted with JetBrains AI Assistant: what it rewards, where it breaks, and how to keep the workflow honest.
-
Windsurf: Flows as a new way to carry intent across files
A field-tested take on flows as a new way to carry intent across files with Windsurf: what it rewards, where it breaks, and how to keep the workflow honest.
-
Cursor: Writing a .cursorrules file that did not embarrass us
A field-tested take on writing a .cursorrules file that did not embarrass us with Cursor: what it rewards, where it breaks, and how to keep the workflow honest.
-
Google Gemini: Gemini Code Assist versus the API, a first year comparison
A field-tested take on Gemini Code Assist versus the API, a first year comparison with Google Gemini: what it rewards, where it breaks, and how to keep the workflow honest.
-
ChatGPT Codex: Writing AGENTS.md for a team that had never used agents
A field-tested take on writing AGENTS.md for a team that had never used agents with ChatGPT Codex: what it rewards, where it breaks, and how to keep the workflow honest.
-
Claude Code: Writing your first CLAUDE.md and learning what to leave out
A field-tested take on writing your first CLAUDE.md and learning what to leave out with Claude Code: what it rewards, where it breaks, and how to keep the workflow honest.
-
GitHub Copilot: Teaching Copilot your repository one readme at a time
A field-tested take on teaching Copilot your repository one readme at a time with GitHub Copilot: what it rewards, where it breaks, and how to keep the workflow honest.
-
Local LLMs with Ollama: First week running Ollama on a work laptop
A field-tested take on the first week running Ollama on a work laptop with Local LLMs with Ollama: what it rewards, where it breaks, and how to keep the workflow honest.
-
Agentic coding: First overnight agent run we dared to approve
A field-tested take on the first overnight agent run we dared to approve with Agentic coding: what it rewards, where it breaks, and how to keep the workflow honest.
-
MCP servers: First MCP server that earned its keep
A field-tested take on the first MCP server that earned its keep with MCP servers: what it rewards, where it breaks, and how to keep the workflow honest.
-
AI coding workflows: First weekly review that made AI-assisted work sustainable
A field-tested take on the first weekly review that made AI-assisted work sustainable with AI coding workflows: what it rewards, where it breaks, and how to keep the workflow honest.
-
Prompt engineering for code: First prompts we wrote that still pass review a year later
A field-tested take on the first prompts we wrote that still pass review a year later with Prompt engineering for code: what it rewards, where it breaks, and how to keep the workflow honest.
-
Cline and Roo Code: First open-source agent we actually trusted
A field-tested take on the first open-source agent we actually trusted with Cline and Roo Code: what it rewards, where it breaks, and how to keep the workflow honest.
-
JetBrains AI Assistant: AI Assistant day one in a JetBrains shop
A field-tested take on AI Assistant day one in a JetBrains shop with JetBrains AI Assistant: what it rewards, where it breaks, and how to keep the workflow honest.
-
Windsurf: First look at Cascade after a year of Cursor
A field-tested take on the first look at Cascade after a year of Cursor with Windsurf: what it rewards, where it breaks, and how to keep the workflow honest.
-
Cursor: First hour inside Cursor after years of VS Code
A field-tested take on the first hour inside Cursor after years of VS Code with Cursor: what it rewards, where it breaks, and how to keep the workflow honest.
-
Google Gemini: First million-token prompt we actually sent
A field-tested take on the first million-token prompt we actually sent with Google Gemini: what it rewards, where it breaks, and how to keep the workflow honest.
-
ChatGPT Codex: First Codex task that opened a real pull request
A field-tested take on the first Codex task that opened a real pull request with ChatGPT Codex: what it rewards, where it breaks, and how to keep the workflow honest.
-
Claude Code: First week of Claude Code at the terminal
A field-tested take on the first week of Claude Code at the terminal with Claude Code: what it rewards, where it breaks, and how to keep the workflow honest.
-
GitHub Copilot: First honest week with GitHub Copilot on a real codebase
A field-tested take on the first honest week with GitHub Copilot on a real codebase with GitHub Copilot: what it rewards, where it breaks, and how to keep the workflow honest.