Nectar FAQ
Short answers to the common questions about Nectar.
Nectar FAQ
Short answers to the common questions about Nectar.
Related:
#What is Nectar in one sentence?
A memory layer that gives every file a stable identity and a plain-language description of what it does, so your AI coding assistant can find code by meaning instead of by file name.
#Do I need to change how I write code?
No. No special comments, markers, naming conventions, or metadata. You write, name, and organize code as you do today, and the descriptions are produced for you automatically. Your editor, version control, and build are untouched.
#Does it modify my source files?
Never. It only reads your source. The single file it writes is .honeycomb/nectars.json at the project root, kept separate from your code and fully regenerable.
#What does the first scan cost, and is it recurring?
It is a one-time cost per project. About 200 files costs roughly thirty cents, 2,000 files about three dollars, and 10,000 files about fifteen dollars. Preview the exact figure with nectar brood --dry-run, which spends nothing. Ongoing cost is minimal, because only meaningfully changed files are re-described.
#What happens when I rename or move a file?
The description follows the file. Identity is a stable tag independent of name and location, so renames, moves, and full directory reorganizations keep every description intact, with no re-describe cost.
#How do my teammates get the understanding?
Commit .honeycomb/nectars.json. A teammate who clones the repo inherits every description instantly, for free, with no re-scan, and it works offline.
#Does it re-scan on every edit?
No. Edits are debounced, so it waits for a pause, and a file is re-described only when its content meaningfully changed. Cosmetic reformatting is ignored.
#Does my code leave my machine?
Only a file's contents are sent, per file, to the description model, and only through a gateway you configure yourself. Only the generated descriptions come back, and identity is created and stored locally. If no source may leave your network, that is a matter of how you configure the gateway.
#Does it replace my editor's search or "go to definition"?
No, they complement each other. Editor search and symbol navigation are structural, for when you know the exact name. Nectar is semantic, for when you know the idea but not the name.
#What if a description looks wrong?
It usually corrects itself on the next meaningful edit, or you can force a fresh description with nectar brood --force.
#What happens if I stop using Nectar?
Nothing is lost. The map is a committed, readable, portable file that does not depend on a running service or subscription. Because your source was never modified, there is nothing to clean up.
#Which assistants work with it?
The ones the Apiary supports: Claude Code, Cursor, and Codex today, with Hermes, pi, and OpenClaw in progress. Nectar feeds them through Honeycomb's shared memory rather than replacing them.