Implementations¶
oryxkb¶
oryxkb is the first implementation. It is an open-source Python tool. It reads
the oryxkb/ folder, checks it (oryxkb check, with --strict for CI), renders cards for agents, tracks
changes between builds, and resolves citations. The JSON Schema in the spec's schema/ folder is generated from its
models for now. oryxkb additionally accepts an outputs map for pipeline steps that save several tables
(documented in the oryxkb docs).
Try it. Install steps are in the oryxkb docs. Then, from a project folder:
oryxkb check oryxkb/project.yaml
A file with one typo reports:
ERROR oryxkb/project.yaml:34: datasets.Rents.columns.rent_b: unknown key 'descripton'
1 datasets; 1 warnings (1 ERROR)
A walk-through: Write your first file.
Where the spec and the tool disagree, the difference is a bug in one of them. Report it. Which one wins is not settled while the spec is a draft.
Not built yet in oryxkb: telling the agent to compare the grains of both sides before a join (the card shows each grain and the declared joins).
Others¶
Services and other tools that implement the spec will be listed here. To be listed, an implementation reads
version 1 files and treats unknown x- keys as allowed.