% ============================================================================
\section{Conclusion}
\label{sec:conclusion}
% ============================================================================

This paper has described VQuant, a deployed platform that turns a natural-language request into
a quantitative financial report under one organising principle: the language model plans and
synthesises, and deterministic code retrieves data and performs computation. We specified four
mechanisms as implemented and audited against the source code---a recursive, batched, streaming
tool-calling loop over a 245-tool catalogue; a three-stage context-management procedure with its
actual constants; a single-call computation convention and the process-isolated sandbox that
executes it; and a document pipeline from one Markdown artifact to PDF, DOCX, Beamer, and
live-formula spreadsheets---and we corrected the earlier description of the system where the
code differs from it. In the vocabulary of design science, the paper contributes a situated
implementation and a demonstration; the evaluation step is specified but not performed.

The limits of what has been shown are therefore clear. The architecture removes numbers from
the model's remit, but whether this reduces unsupported numbers to near zero, by how much the
single-call convention reduces round trips and at what cost in failed executions, how much the
summarisation procedure loses, and what the design costs relative to simpler agents are empirical
questions that the protocol of Section~\ref{subsec:protocol} is written to answer and that this
paper does not. Executing that protocol is the natural next step and the one on which any claim
of superiority must wait. Three engineering changes are also warranted by the audit: a bound on
recursion depth, a per-model price table in the ledger, and the restoration of the pre-built
numerical routines to the tool catalogue. Beyond these, the design admits extensions that the
protocol could also evaluate: a verification pass in which a second model audits the first's
narrative against the computed numbers, provider-independent tool exposure through the Model
Context Protocol, and generalisation of the deterministic-computation substrate to other domains
in which auditable numbers matter. We offer the specification, the audit, and the protocol as the
contribution, and the pattern they describe---a language model as a disciplined orchestrator of
trustworthy tools rather than as an unverified oracle---as a hypothesis to be tested.