# Trader Tool

**Today: a trade journal with data review before saving.
Our vision: a trader's workspace linking the trader's own plan, market context,
and reflection on results.**

## First audience and problem

Our initial segment is a **hypothesis**: self-directed traders who save trade
screenshots, use Telegram, and postpone manual journaling. Saving a screenshot
is easy; copying the numbers, checking a record, and returning to it is another
task. Decision rationale and results can remain scattered.

## Value hypothesis

Get to a confirmed journal entry with less manual input, then return to recent
entries for reflection. We do not claim uniqueness or proven superiority over
a spreadsheet or another trading journal. The future workflow is:
**market context → my plan → trade or skip → journal → review**.

## Today: early prototype

The prototype centres on a journal in a Telegram Mini App and bot. The demo
shows fields from a synthetic screenshot, user review and corrections, explicit
confirmation, persistence and viewing, followed by a mocked assistant response.
Unknown trade direction is not automatically treated as long.

**All demo records are synthetic. Model responses are mocked and clearly
labelled as simulation.** This demonstrates the application workflow, not live
Claude quality or a customer launch.

## Claude's intended role

An integration path is prepared for extracting visible screenshot fields and
asking questions about recent confirmed entries. Users review before saving.
Avoiding trading commands is a requirement for the assistant's role, not a
validated guarantee about model behaviour.

## Vision: context beside a decision

Analysis is a separate research project for market data, context, and forecasting
hypotheses. The proposed first connection saves a timestamped Analysis snapshot
beside a user's plan and journal entry, so expectations, actions, and results
could later be compared. This connection is not implemented. Forecasting remains
a research goal; accuracy, usefulness, and profitability are unproven.

## Demand and business model

A future pilot would compare logging with each participant's usual method,
measure time and corrections, and check whether they return for another review.
That pilot has not happened and is not required to finish this presentation.
The business model is a **subscription hypothesis**: a self-directed trader pays
for journaling and regular review. Pricing and willingness to pay are unvalidated.

**The trader remains responsible. No trading signals or automated execution. DYOR.**