CodePlan: Repository-level Coding using LLMs and Planning · HackerPapers discussions

CodePlan: Repository-level Coding using LLMs and Planning — HN submission

Fri, 10 May 2024 16:03:07 GMT

Show HN: I built an AI agent that upgrades NPM packages

Hi HN, I built a tool [1] that resolves breaking changes when you upgrade npm packages. I know keeping my dependencies updated is good hygiene but I really have no motivations to do them. The time to fix breaking changes then validate them makes the ROI lopsided. Automating these helps me reduce the barrier to a good habit. It works on typescript and tsx projects. BYOK (bring your own GPT-4 key) for the codegen. MIT license. How does it work? - Bumps the package version, builds your project, and then runs tsc over your project to understand what breaks - Uses ts-morph [2] to create an AST of…

5 points; 2 discussion replies on HN. Paper and source history. A submission is not an endorsement.

CodePlan: Repository-level Coding using LLMs and Planning — HN submission

Mon, 06 May 2024 14:55:41 GMT

Show HN: Can AI keep my code up-to-date?

Hey HN, We are open sourcing (https://github.com/xeol-io/bumpgen/), a tool that uses an LLM to upgrade your dependencies to their newer versions. We are building this for our customers (security teams) who need to upgrade their dependencies for security and compliance but lack engineering resources. Most patching tend to be painless but major version upgrades will cause breaking changes that require engineers to fix. We want bumpgen to reduce this engineering cost to security. It has been an interesting experience using AI to identify breaking changes, fix them, then propagate the fixes…

2 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

CodePlan: Repository-level Coding using LLMs and Planning — HN submission

Tue, 30 Apr 2024 14:35:04 GMT

Show HN: Bumpgen – upgrade NPM packages using AI

Hey HN, we are building bumpgen (https://github.com/xeol-io/bumpgen/) to remove the toil of fixing breaking changes during version bumps. bumpgen bumps your npm package version then generates the fixes to potential breaking changes. There were some interesting challenges we encountered using an LLM to fix breaking changes: [1] Finding the breaking changes → know how is a dependency used through the codebase [2] Knowing how to fix the a breaking change → know how the dependency has changed from one version to another [3] Understanding how the fix has modified existing behavior → know how the…

20 points; 9 discussion replies on HN. Paper and source history. A submission is not an endorsement.

CodePlan: Repository-level Coding using LLMs and Planning — mentioned by famouswaffles

Wed, 27 Sep 2023 22:06:50 GMT

This looks promising when Microsoft release the code for it. https://arxiv.org/abs/2309.12499

Paper and discussion history. A mention is not an endorsement.

CodePlan: Repository-level Coding using LLMs and Planning — mentioned by famouswaffles

Tue, 26 Sep 2023 00:19:57 GMT

>It can't even generate flashcards from a textbook chapter, because it can't load the entire chapter into memory. Heck, it doesn't even know what textbook I'm talking about; I have to provide the content! Okay...? That's a context window problem. and you could manage it if you sent the textbook in chunks. >The improvement GPT 5 would have to provide is multiple orders of magnitude in order for this to be a realistic proposition. No..it wouldn't https://arxiv.org/abs/2309.12499

Paper and discussion history. A mention is not an endorsement.

CodePlan: Repository-level Coding using LLMs and Planning — mentioned by famouswaffles

Mon, 25 Sep 2023 20:08:48 GMT

>And it is not true that computing power will continue to reduce; Moore's Law has been dead for some time now, and if incremental growth in LLMs require exponential growth in computing power the marginal difference won't matter. This is a lot of unfounded assumptions. You don't need Moore's Law. GPU's are not really made with ML training in mind. You don't need exponential growth for anything. The money Open ai spent on GPT-4 a year ago could train a model twice as large today. and that amount is a drop in the bucket for the R&D of large corporations. Microsoft gave open ai 10B. amazon gave…

Paper and discussion history. A mention is not an endorsement.

CodePlan: Repository-level Coding using LLMs and Planning — mentioned by famouswaffles

Mon, 25 Sep 2023 19:06:42 GMT

https://arxiv.org/abs/2309.12499

Paper and discussion history. A mention is not an endorsement.

CodePlan: Repository-level Coding using LLMs and Planning — HN submission

Mon, 25 Sep 2023 14:45:57 GMT

CodePlan: Repository-level Coding using LLMs and Planning

3 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.