Agents In Use
What’s new todayLast update
108 use cases added today

Use Claude to reconstruct crypto acquisition records

Matt Saucedo used Claude Code to investigate a discrepancy between a crypto exchange's 1099 and his original trade records. The exchange had treated a transfer back from an external wallet as the acquisition date, while the holdings had actually been acquired years earlier.

Source published mattsaucedo.substack.comAgent: Claude
OUR ADAPTATION

Suggested prompt

Suggested prompt written by Agents In Use, not the source author. Replace bracketed inputs and review tool permissions first.

Compare my exchange 1099 with original trades and wallet transfers. Flag inconsistent acquisition dates, cite primary authority, and write a lot-level memo before drafting adjustments.

Evidence at a glance

Sourced account
Source author
Matt Saucedo
Original account
Read the source ↗
Published
Source link checked
Reported result
A cited basis-reconstruction memo documenting lots and original acquisition data, followed by Form 8949 adjustment entries for the broker-date discrepancy.
Human involvement
Saucedo could inspect and correct files; no separate approval checkpoint for these entries is described.

A linked account supports attribution. It does not mean we independently reproduced its result.

Report an issue with this caseOpen correction email

Opens your email app with the case link included. Nothing is sent automatically.

How the agent did it

  1. Supply the broker 1099 and original trade records.
  2. Compare acquisition dates and research primary authority.
  3. Write a lot-level reconstruction memo and linked form adjustments.

Matt Saucedo used Claude Code to investigate a discrepancy between a crypto exchange's 1099 and his original trade records. The exchange had treated a transfer back from an external wallet as the acquisition date, while the holdings had actually been acquired years earlier.

The agent compared the supplied history with the broker document, identified inconsistent dates and researched statutory authority. It wrote a basis reconstruction memo recording individual lots, original cost basis and the authority for overriding broker dates. Saucedo reports that it then prepared Form 8949 entries with adjustment codes.

This work sat inside a project with separate folders for research, computation worksheets and final PDFs. Before calculations, Claude built reference files from primary materials such as IRS publications and statute text. Worksheets linked back to those references, including the holding-period authority used in the crypto analysis. A root CLAUDE.md file supplied operating rules and tracked open questions across sessions.

Saucedo describes the reference files as material he could read, verify and correct. The result was a documented reconstruction and form entries based on supplied records. The article does not describe a separate human approval checkpoint for those particular entries, so their acceptance should not be assumed.

Prompt
Suggested prompt (written by Agents In Use, not the author's)

Source: mattsaucedo.substack.com · Matt Saucedo

Frequently asked questions

What inconsistency triggered the work?

The broker used the wallet-transfer date, while the original records showed purchases years earlier.

What made the result inspectable?

The memo recorded individual lots, original basis and authority, linked to a local research library.

Did calculations come first?

Saucedo describes building a primary-source knowledge base before using it in computation worksheets.

More details

About this account

Claude compared records, researched authority and wrote a reconstruction memo and Form 8949 entries.

Firsthand source opened and read. Suggested prompt written by us from the reported workflow; no original task prompt is quoted.

Credit: Matt Saucedo

Source published

Added to Agents In Use

Source checked

Tools and topics

Claude Code · CLAUDE.md · Trade records · Local Markdown

Crypto records · Cost basis · Form 8949

Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.