菜单
Use when interacting with dodder ZettelIds (two-part IDs like `ceroplastes/midtown`), seeding a new repo via `dodder init -yin` / `-yang`, choosing predictable or random ID allocation, or interpreting "zettel ids exhausted" errors at the command line.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Guides day-to-day dodder and der operations including creating zettels, querying and filtering objects, checking out working copies, organizing content, syncing with remotes, and managing tags, types, and dormant objects. Activated by requests to create, edit, organize, query, show, checkout, checkin, push, pull, sync, tag, or otherwise operate on dodder repositories.
Use when working with object genres, adding genre filters to queries, parsing sigils like :z :t :e :k :b, or modifying genre-based dispatch. Also applies when encountering Genre byte type, genre bitwise operations, MakeGenre, or optimizedQueries maps.
Introduces dodder, a distributed zettelkasten and content-addressable blob store. Covers core concepts (zettels, object IDs, tags, types, blobs, the store), installation via Nix, a first-steps walkthrough including repository initialization and zettel creation, and the mental model for working with dodder. Activated when a user is new to dodder, asks what dodder is, wants to get started, install, set up, or learn how dodder works.
Use when working on the ZettelId internals
Use when working on the dodder/madder blob-store internals
Use when creating, querying, or sharing dodder blob stores at the command line
| name | features-zettel_ids |
| description | Use when interacting with dodder ZettelIds (two-part IDs like `ceroplastes/midtown`), seeding a new repo via `dodder init -yin` / `-yang`, choosing predictable or random ID allocation, or interpreting "zettel ids exhausted" errors at the command line. |
Dodder assigns each zettel a unique two-part string identifier (e.g.
ceroplastes/midtown). Each repo has a finite pool of available IDs, fixed
at dodder init time; once that pool is empty, no more zettels can be
created in that repo.
A ZettelId has the form left/right, where both halves are non-empty
identifiers. The left and right halves are drawn from two word lists called
Yin (left side) and Yang (right side). The same word list pair is
used for the lifetime of the repo.
At dodder init time the Yin and Yang word lists are copied from files
passed via the -yin and -yang flags. The total ID space for the repo
equals len(Yin) * len(Yang) --- two 200-word lists give 40 000 possible
zettel IDs.
dodder init \
-yin path/to/yin.txt \
-yang path/to/yang.txt \
...
The lists are immutable after init: once a repo is created, you cannot swap in a different Yin/Yang to enlarge the pool. Pick lists large enough for the lifetime of the repo (or accept that you may need to spin up a new repo when you run out).
Two modes, controlled by a CLI config flag:
When the pool is empty, dodder new (and anything else that mints zettels)
fails with zettel ids exhausted. There is no in-place expansion path ---
remediation is repo-level (start a new repo with a larger Yin/Yang, or
archive existing zettels into a different store).