Workflowrun again on a new scope

Documentation mapping to schema

An operator's documents can have no path to a queryable record

The problem to solve

An operator holds what they know in documents. The facts are on the page, in prose and tables and appendices, and nothing downstream can query a page. Someone re-keys them, or nobody does.

The solution it delivers

Source documents become schema-conforming records an operator can query, checked by an independent critic before a person reads them, and every correction the operator makes teaches the next render.

01

The run as one route: which tools get involved, in what order, and where the chain is not mechanised yet.

DOCS, NO QUERY PATHOPERATOR CORRECTSaccept, edit, or rejectSCRIBEProfile, propose, critic, ren…QUERYABLE RECORDSAn operator's documents become schema-conforming records their systems can actually query.

The scattered mark is the problem the run starts from. Every tool sits off the route it drives, glyph and name, and opens its own page. A diamond is a decision the work cannot pass without today, with no tool yet doing it. The accented mark near the end is the named solution, into Solutions; the line beneath it is the outcome the run closes on.

02Which tools, in what capacity
actuate
The whole loop, profile, propose, critic, render
substrate
Model calls
03How the work crosses between them

Source docs → profile is internal to Scribe, not a cross-tool handoff; the operator correction loop is a signal append (accept/edit/reject + rationale + embedding, into the tenant-gated Postgres correction store); corpus learning is a contract read (retrieval reads the corpus's own declared state at the next render, nothing is sent).

The tools this workflow runs through