lukeundtrug

3 ms·

1. 1.▲
   by lukeundtrug 5mo ago

   My guess is that we underestimate how much non-Java data and context in general is needed to create a good Java coding model. It could be true that a good Java model would be of 80-90% the size of a comparable overall coding model. Obvious
2. 2.▲
   by lukeundtrug 5mo ago

   That sounds great, thanks for sharing your thoughts! It sure sounds like we have similar things in mind. I basically try to build the proper graph representation of the code during runtime, so all caller/callee relationships plus type
3. 3.▲
   by lukeundtrug 5mo ago

   I definitely share the same sentiment. I don’t want to overload the llm with many tools. Better to have a few opinionated and flexible ones, but yeah, keeping the balance is hard. I would say the main two tools are get-symbol-context and ge
4. 4.▲
   by lukeundtrug 5mo ago

   Happened to have written both a tool and a blog post about the topic. It’s more about the different technical approaches you have in solving the problem but it might still interest you :) https://www.context-master.dev/blog&
5. 5.▲
   by lukeundtrug 7mo ago

   Thanks for taking the time to check it out and for the kind words! I really appreciate it. I totally get sticking with your current approach. Your workflow sounds very intriguing as well. A combination of both approaches might really be ver
6. 6.▲
   by lukeundtrug 7mo ago

   I built myself an AST based solution for that during the last 6 months roughly. I always wondered whether grep and agent-based discovery will be the end of it and thought it just has to be better with a more deterministic approach. In the e