Skip to content
Get packs

Financial Data Cross-Check

Pull company financials from two independent sources and flag discrepancies over 1%.

Source Download skill
Repository
xbtlin/ai-berkshire
Stars
16,982
Licence
MIT
Files
1 file

What it does

This skill gives an AI agent a strict workflow for gathering company financial figures such as revenue, net profit, margins and cash flow. Each key number must be fetched from two independent sites, compared, and marked as consistent, showing a discrepancy, or a major discrepancy needing check against the original filing. It also covers stock price adjustment rules, Taiwan-stock data via a bundled FinMind script, and common reasons numbers differ (GAAP vs non-GAAP, exchange rates, fiscal years). The agent is told to check the current date first and state the data cutoff in the report.

Great for

  • Investment analysts who want numbers verified across two sources before relying on them
  • Researchers writing company reports on US-listed, Hong Kong, A-share or Taiwan-listed firms
  • Developers building a repeatable financial-research workflow in the AI Berkshire repository

Skip it if

  • Investment decisions without independent review: the skill is a research process, not financial advice, and the sources it names can be wrong or stale
  • Private companies or sources that are not on the list: the skill gives no method for them beyond marking figures as estimates
  • Real-time trading data: it relies on sites that may lag behind the latest filings

Try asking

  • “Compare PDD's latest annual revenue and net profit across macrotrends and stockanalysis and flag any gap over 1%”
  • “Pull TSMC's last five years of core financials and monthly revenue using the Taiwan data tool”
  • “Build a cross-checked valuation table for Tencent using AASTOCKS and TCEHY data”
Step-by-step guide: how to use Financial Data Cross-CheckInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the MIT licence and an attribution note.

    financial-data.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip financial-data.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip financial-data.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Ask for the data cutoff date in the report header, since the skill requires checking today's date first
  • Store any FinMind token in the environment variable or the git-ignored local file, never in reports or commits
  • Check the original 10-K or annual report whenever the two sources disagree by more than 5%

What's inside

  • SKILL.md

Redistributed unmodified from xbtlin/ai-berkshire under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.