May 2025
31 published posts filed under May 2025. Each article shows its actual publication date.
-
Cline and Roo Code: What closed-source agents still did better, and where they did not
A field-tested take on what closed-source agents still did better, and where they did not with Cline and Roo Code: what it rewards, where it breaks, and how to keep the workflow honest.
-
JetBrains AI Assistant: Teaching Junie through existing Markdown docs
A field-tested take on teaching Junie through existing Markdown docs with JetBrains AI Assistant: what it rewards, where it breaks, and how to keep the workflow honest.
-
Windsurf: Write mode restraint in the first month of use
A field-tested take on write mode restraint in the first month of use with Windsurf: what it rewards, where it breaks, and how to keep the workflow honest.
-
Cursor: Learning to write intent in comments, not in chat
A field-tested take on learning to write intent in comments, not in chat with Cursor: what it rewards, where it breaks, and how to keep the workflow honest.
-
Google Gemini: Honest cost of long-context prompts in a small team
A field-tested take on the honest cost of long-context prompts in a small team with Google Gemini: what it rewards, where it breaks, and how to keep the workflow honest.
-
ChatGPT Codex: Small playbook that made Codex predictable
A field-tested take on the small playbook that made Codex predictable with ChatGPT Codex: what it rewards, where it breaks, and how to keep the workflow honest.
-
Claude Code: Moment Claude Code beat a human reviewer on taste
A field-tested take on the moment Claude Code beat a human reviewer on taste with Claude Code: what it rewards, where it breaks, and how to keep the workflow honest.
-
GitHub Copilot: First model switch inside Copilot Chat, and what it taught us
A field-tested take on the first model switch inside Copilot Chat, and what it taught us with GitHub Copilot: what it rewards, where it breaks, and how to keep the workflow honest.
-
Local LLMs with Ollama: Air-gapped development as a serious practice, not a demo
A field-tested take on air-gapped development as a serious practice, not a demo with Local LLMs with Ollama: what it rewards, where it breaks, and how to keep the workflow honest.
-
Agentic coding: First PR an agent opened that we merged without edits
A field-tested take on the first PR an agent opened that we merged without edits with Agentic coding: what it rewards, where it breaks, and how to keep the workflow honest.
-
MCP servers: What the protocol gave us and what our product still had to
A field-tested take on what the protocol gave us and what our product still had to with MCP servers: what it rewards, where it breaks, and how to keep the workflow honest.
-
AI coding workflows: First retrospective on what AI did and did not change
A field-tested take on a first retrospective on what AI did and did not change with AI coding workflows: what it rewards, where it breaks, and how to keep the workflow honest.
-
Prompt engineering for code: First prompt rot we detected, and the lessons from it
A field-tested take on the first prompt rot we detected, and the lessons from it with Prompt engineering for code: what it rewards, where it breaks, and how to keep the workflow honest.
-
Cline and Roo Code: Quiet pleasure of reading every tool call
A field-tested take on the quiet pleasure of reading every tool call with Cline and Roo Code: what it rewards, where it breaks, and how to keep the workflow honest.
-
JetBrains AI Assistant: Early taste tests: JetBrains AI versus the newcomers
A field-tested take on early taste tests: JetBrains AI versus the newcomers with JetBrains AI Assistant: what it rewards, where it breaks, and how to keep the workflow honest.
-
Windsurf: Windsurf setting that changed how we reviewed code
A field-tested take on the Windsurf setting that changed how we reviewed code with Windsurf: what it rewards, where it breaks, and how to keep the workflow honest.
-
Cursor: Pricing tier choices with incomplete information
A field-tested take on pricing tier choices with incomplete information with Cursor: what it rewards, where it breaks, and how to keep the workflow honest.
-
Google Gemini: First real comparison we ran between Gemini and the rest
A field-tested take on the first real comparison we ran between Gemini and the rest with Google Gemini: what it rewards, where it breaks, and how to keep the workflow honest.
-
ChatGPT Codex: Choosing Codex for upgrades, humans for design
A field-tested take on choosing Codex for upgrades, humans for design with ChatGPT Codex: what it rewards, where it breaks, and how to keep the workflow honest.
-
Claude Code: Slash commands as the first real team workflow
A field-tested take on slash commands as the first real team workflow with Claude Code: what it rewards, where it breaks, and how to keep the workflow honest.
-
GitHub Copilot: When to lean on Copilot CLI instead of the IDE extension
A field-tested take on when to lean on Copilot CLI instead of the IDE extension with GitHub Copilot: what it rewards, where it breaks, and how to keep the workflow honest.
-
Local LLMs with Ollama: Pairing Ollama with Continue for a first private IDE setup
A field-tested take on pairing Ollama with Continue for a first private IDE setup with Local LLMs with Ollama: what it rewards, where it breaks, and how to keep the workflow honest.
-
Agentic coding: Cost-per-merged-change metric we invented out of necessity
A field-tested take on the cost-per-merged-change metric we invented out of necessity with Agentic coding: what it rewards, where it breaks, and how to keep the workflow honest.
-
MCP servers: First MCP server a security review actually approved
A field-tested take on the first MCP server a security review actually approved with MCP servers: what it rewards, where it breaks, and how to keep the workflow honest.
-
AI coding workflows: Diff as primary artefact, chat as exhaust, a first lesson
A field-tested take on the diff as primary artefact, chat as exhaust, a first lesson with AI coding workflows: what it rewards, where it breaks, and how to keep the workflow honest.
-
Prompt engineering for code: How we measured prompt quality before we had metrics
A field-tested take on how we measured prompt quality before we had metrics with Prompt engineering for code: what it rewards, where it breaks, and how to keep the workflow honest.
-
Cline and Roo Code: First month of honest logs from an open-source agent
A field-tested take on a first month of honest logs from an open-source agent with Cline and Roo Code: what it rewards, where it breaks, and how to keep the workflow honest.
-
JetBrains AI Assistant: First CI run with Junie as a contributor
A field-tested take on the first CI run with Junie as a contributor with JetBrains AI Assistant: what it rewards, where it breaks, and how to keep the workflow honest.
-
Windsurf: First MCP server we plugged into Windsurf
A field-tested take on the first MCP server we plugged into Windsurf with Windsurf: what it rewards, where it breaks, and how to keep the workflow honest.
-
Cursor: Undoing Cursor changes on purpose, a first discipline
A field-tested take on undoing Cursor changes on purpose, a first discipline with Cursor: what it rewards, where it breaks, and how to keep the workflow honest.
-
Google Gemini: When to ask Gemini and when to open the logs yourself
A field-tested take on when to ask Gemini and when to open the logs yourself with Google Gemini: what it rewards, where it breaks, and how to keep the workflow honest.