Skip to content

Latest commit

 

History

History
31 lines (24 loc) · 1.75 KB

File metadata and controls

31 lines (24 loc) · 1.75 KB

Dispositions, reason tokens, and health codes

The one-word outcomes docket reports, the reason tokens a blocked finalize prints, and the health codes a status scan raises are each defined in exactly one owning surface. This page points at those owners and copies none of the vocabularies, so it can never list a token the current binary no longer emits.

Dispositions

A disposition is the one-word outcome an operation reports: applied, no-op, refused, or error. This closed vocabulary is owned by the docket-convention skill's Step-0 preamble (../../skills/docket-convention/SKILL.md), which states it as "the closed vocabulary applied | no-op | refused | error" and defines what each one obliges a caller to do. The exact closed set the binary emits is also machine-readable via docket schema (the schema/vocabulary command).

Reason tokens

When finalize refuses or halts, it names a typed reason token (for example, a mismatched PR head or an unresolved review state). The token vocabulary and each token's remedy are owned by the finalize skill's failure reference, ../../skills/docket-finalize-change/references/gate-failure.md. Read the current token and its recovery there, not from memory.

Health codes

A health check is a status-time scan for things a human should look at: stale claims, broken links, stalled dependencies. The health codes a scan can raise, and their current data, are owned by docket status --json — its output carries the live health section for your repo. The human-readable form is docket status. The convention skill's lifecycle and learnings-ledger sections give the background each code is checking against.