Nectar
Nectar reads every file in your project, writes a plain-language description, and gives your AI coding assistant a way to find code by meaning, not file name.
Nectar is the understanding layer in the Apiary. It reads every file in your project, gives it a stable identity, and writes a plain-language description of what it does, so your AI assistant can find code by meaning instead of guessing from file names.
Nectar is the understanding layer of the Apiary. It reads every file in your project once, gives each one a stable identity, and writes a short plain-language description of what it does. Your AI assistant then searches those descriptions instead of guessing from file names, and finds the right code the first time.
#What problem does nectar solve?
AI coding assistants are good at reading code and bad at finding the right code to read. Ask "where do we handle logins" and most assistants go hunting for a file named login.ts or auth.ts. In a real codebase, that logic often lives somewhere less obvious, like session-refresh.ts, three folders deep, with a name that gives no hint of what it does.
When the assistant can't find the right file, it reads the wrong one, gives you a confident answer built on incomplete information, and you end up doing the search yourself anyway. That's the gap nectar closes.
#What does nectar actually do?
Nectar reads each file in your project and writes a short description of what it does in plain language, something like "refreshes login tokens on each authenticated request." It stores that description and keeps it current as your code changes. When you or your assistant later ask about anything related to logins, the search runs against those descriptions, not just file names, so it finds every file that plays a part, not just the ones with the right name.
Think of it as the index at the back of a book, except this index actually read every chapter and knows what each one is about, not just which words show up on the page.
#What do you get from it?
- Search that matches meaning. Ask "where do we handle payments" and get the files that do the work, no matter what they're named.
- Understanding that survives refactors. Each file gets a stable identity, so its description follows it through renames and moves. Reorganizing your folders doesn't reset anything.
- A low, one-time cost. Scanning a typical project of about 2,000 files costs around three dollars. You can preview the exact number before spending anything.
- Free sharing. The understanding is stored in one file you commit to your repo, so teammates who clone the project inherit every description instantly, at no extra cost.
- Your source, untouched. The only file nectar writes is its own map. Not one character of your actual code changes.
#Where does it fit in the Apiary?
Nectar is one of the products in the Apiary. Honeycomb holds your working memory, and nectar makes your codebase searchable by meaning, feeding its results into honeycomb's shared recall. Your assistant benefits automatically, without you opening a separate search tool.
#Common questions
What is nectar in one sentence? A layer that gives every file in your project a stable identity and a plain-language description, so your AI assistant can find code by meaning instead of by file name.
Do I have to change how I write code? No. You write, name, and organize your code exactly as you do today. Nectar works in the background.
Does it read or send my code anywhere unsafe? It only reads your source to write descriptions, and only sends file contents through a gateway you configure. See how nectar works and the nectar FAQ for the full picture.