Command Explorer › scale: — the reply's measure › deep
scale:deep
The full version: thorough treatment, evidence in view.
In the scale: domain
scale:brief — The short version: verdict first, essentials only.scale:deep — The full version: thorough treatment, evidence in view.scale:executive — Bottom line first, decision-ready, as short as the substance allows — the depth is in the work, not the word count.scale:wordy — The long version, fully aired out — expansive on what's known, with the effort capped: no deep dive, no rabbit holes.scale:broad — The broad version: wide coverage over depth — everything touched, nothing exhausted.scale:compact — Short with the substance kept: real work done and packed tight — on the road toward executive, less extreme.What it does
The thorough treatment: the question answered completely, with its evidence in view. Depth names both the work and its showing: the effort behind the reply is scaled to the completeness the reply claims — it measures the reply, and is not a separate instruction about how to think.
The registry's own words: The full version: thorough, complete treatment of the question, with its evidence in view. Depth names both the work and its showing: the effort behind the reply is scaled to the completeness the reply claims — it measures the reply, and is not a separate instruction about how to think.
Fire it — usage examples
“Is this lease clause actually enforceable?
scale:deep”→ the full treatment: the clause, the reasoning, the evidence, the uncertainties marked as uncertainties.
Your agent answered in three sentences and you want the whole picture. Reply to it:
re:de→ the question reopened in full — reasoning, evidence, and the open questions behind the question.
“Should we open a second location this year?
scale:deep/ci”→ the thorough treatment, with the sources behind its load-bearing claims shown inline — depth and evidence stacked in one token.
Said aloud: “Lingo, reply deep — the lease clause.”
→ the full treatment, evidence in view.
Spoken. Inside a marked span opened by the wake word “Lingo”, the spoken form is “reply deep”. Dictated directives fire deliberately, never by accident.
What the agent actually does
The agent scopes the question fully — the question asked and the obvious ones behind it — gathers evidence before concluding, and shows it: sources where they were used, reasoning where it isn't obvious.
The baseline procedure, its edge cases, and this directive's full guide live in the implementation guide: implementation/IMPLEMENTATION.md#cmd-scale-deep ↗.
When the host can't
No host capability required. Live retrieval strengthens the evidence but is not required; without it, give the thorough treatment from existing knowledge and mark it as not freshly sourced. Never fabricate citations.
Conformance — “done when”
The reply is a complete treatment with its evidence shown (or the absence of fresh sourcing disclosed).
Persistence class: none. Execution report in the reply: not required. A user can test any of this by hand — fire the directive, then check the result against the store, the schedule, or the reply itself.