Skip to main content

Prepare the future source-to-draft workflow

For review — no decision recorded

Prepared 5 October 2026.

Decision requested​

When automation is selected, which official sources, markets and publication minimums will prove the reduced welcome-offer workflow is useful?

The reduced packages were selected for planning on 14 September 2026. Their start timing remains open. This brief asks about readiness and scope; it does not reactivate the earlier broader automation plan.

Current agreement​

Manual catalogue work can proceed without a parser. Future extraction returns retained facts and sources; editorial preparation creates supported proposals; human review controls publication. These are different responsibilities.

The selected future increment covers one uniquely identified welcome package, including supported repeated deposit stages and distinct cash/spin conditions. Preparation produces structured terms, a factual headline and a short summary. The operator workflow reviews the whole draft and resets approval on every edit.

Proposal for discussion​

Retain that reduced scope for its first future increment. Nominate representative official sources and explicit market contexts, confirm the required term coverage and test whether standard staff forms make nested editing and source inspection usable. Missing or ambiguous packages should produce an actionable gap rather than a guessed card.

A newer eligible proposal should create a separate replacement draft. Operators retain earlier manual work, deliberately reuse suitable text and review the whole replacement. The existing public revision stays selected until another revision is explicitly approved and published.

Illustrative result​

A cash-plus-spins welcome package has different wagering bases for cash and spin winnings. The parser retains them separately; the short copy does not collapse them into one multiplier. If required spin terms are missing, the operator cannot publish a convenient cash-only fragment as though it were the complete package. A supported manual correction keeps its source and review history.

Choices and tradeoffs​

  • Nominate sources and markets that cover each selected field within the planned three-site, up-to-twelve-page parser corpus. An inaccessible source cannot pass as an empty successful example.
  • Agree type-specific publication minimums and which optional unknowns can remain gaps. Stricter requirements may require different source selection.
  • Confirm standard forms are usable for repeated stages and source support. A bespoke nested editor requires a scope and effort decision.
  • Confirm that one wording review round is sufficient for the smaller headline and summary output. Engineering and support defects still require correction.
  • Select the separately needed consumer integration, public-page work and connected rehearsal when scheduling a complete source-to-reader journey.

The reduced approach postpones general profile extraction, additional mechanics, internal analyses, per-item approval and automatic text merging. Those features are not cancelled, delivered or included for free by the smaller estimate.

Effort context​

The 14 September 2026 estimates are active senior-engineer hours, including AI direction, review and corrections: parser 163 / 257 / 399; preparation 50 / 81 / 126; publication 52 / 83 / 126, in optimistic / realistic / pessimistic order. They exclude consumer integration, website adaptation, deployment and the live cross-application rehearsal. Waiting and agent runtime are not active engineering effort. These are dated estimates, not a schedule.

Response requested​

Confirm whether the reduced welcome scope remains the intended first automation increment. Identify nominated source/market cases, required terms and any standard-form concern. State which separate integration outcomes need selection before committing to a connected result. See source to draft and review and publication.

Recorded outcome​

No decision recorded. This brief does not authorize implementation or change delivery timing. Share the specific page and version through the existing discussion channel. Respond with accept, requested changes or defer, identifying the question concerned. How to review explains how an actual agreement is recorded and reflected in the product documents.