Programming · HN · 🌐 International · via agentusecases.ai
Team turns on the strictest compiler settings on a 1.5 million line system and uses Claude to fix what comes up
A developer explained on Hacker News how Claude tackles whole classes of compiler errors, with the team's own skills and scripts, in a huge .NET codebase.
- Profession
- Developer
- Tools
- Claude
- Source
- news.ycombinator.com ↗
- Country
- 🌐 International
- Found via
- agentusecases.ai (Alex Finn) ↗
- Report date
- November 2, 2026
- Verified on
- November 3, 2026
See more: AI for developers

What they did
The team maintains a system of about 1.5 million lines in .NET, in a monorepo with around 50 projects. They set the compiler to treat warnings as errors and be as strict as possible, which produces tens of thousands of items to fix. Instead of going item by item in the IDE, Claude takes one whole class of error at a time and converts the code to the cleaner version, using skills the team built or writing scripts to make the change in bulk. They split the work into three groups: business logic changes, performance changes and cosmetic changes, and leave business logic for dedicated sessions.
How it went
According to the author, over the past year the team found and fixed thousands of bugs, security holes and circular dependencies, increased test coverage and moved all the code to .NET 10 LTS. He says changing more than 200,000 lines this way would be nearly impossible without AI, and that the switch doesn't disrupt colleagues, who no longer use the IDE as their main tool.
Good to know
This is one person's account, with no independent numbers. The author himself says it depends on a strongly typed language and a good compiler, and that they first broke huge files into smaller ones.
The prompt
I turned on the rule [rule name or compiler warning] in the [language and framework] project and it produced [number] occurrences. Group the occurrences by pattern, propose the fix for each pattern and tell me which ones can be done in bulk with a script and which need case by case review. Don't change any business logic: if a fix could change how the system behaves, list it separately and don't apply it. Show me the plan before touching any file.
Replace what is in [brackets]. Prompt written by us from the account, not a literal copy of the source. Do not paste other people’s personal data (privacy laws such as LGPD and GDPR apply).
Summary in our own words. Results as stated in the original account; we have not tested or audited them. Read the source ↗ Case found via agentusecases.ai (Alex Finn); text and prompt rewritten by us. Found a mistake?
Similar cases
01A requirements doc shelved for a year becomes a live website in one day with Claude Code
A tech consultant pulled out the spec for an aggregator of his own posts and shipped the site in a single day with Claude Code.
02Converting Map Data with AI's Help
Simon Willison used Claude and ChatGPT to convert a shapefile of a park boundary into GeoJSON.
03Livelo engineers report up to 50% higher productivity with AI agents
The Brazilian loyalty program brought Kiro and Amazon Q Developer to its team of about 240 engineers, with metrics and governance from day one.
04A leaner MCP server finds wasted AWS spend in fewer calls
A cloud architect rebuilt a 15-tool MCP server with just 4 tools, and the agent now answers where an AWS account is wasting money in 2 calls instead of 7.
05Letting the agent deploy and read production errors on its own
The founder of a deployment platform explains how they designed an MCP server to take the human out of "copy and paste the log".
06Launching a habit app solo and doing the marketing with AI
A dev created Kada, a habit app with no account and no ads, and used AI for marketing.
07Building an "internet TV" in 6 days with a team of agents
A dev created Crom TV, with 24-hour channels of YouTube videos, coordinating several AI agents.
08Asking AI for a specific algorithm for a research project
A developer skeptical of LLMs used Gemini to generate random quantum circuits in the QCOP project.