AI Skill MartBrowse skillsPromptsProgramsLearnWhat is a Skill?
 
← Back to list

Huashu Excel Data Analysis Master

An eight-step data-analysis workflow that profiles and cleans messy Excel/CSV data, pins down metric definitions, reconciles numbers, and has a separate agent stress-test the conclusions.

Data & AnalyticsIntermediate★ 456⑂ 71AI score 10/10Last updated: Nov 4, 2026

What it does

  • Enforces a standard operating procedure: profile → clean → align with the user (+ external benchmarks) → analyze → reconcile → deliver → verify visuals → independent review.
  • Ships scripts (profile_table.py, clean_table.py, scan_traps.py, verify_numbers.py) that mechanically check row-count conservation, sum conservation, and traceability back to source cells.
  • Requires a standalone metric-definition statement (gross vs net, refund handling, de-duplication) so every figure can be challenged rather than merely believed.
  • Writes HTML reports by hand from an explicit design plan (color/type/numerals/space/layout tokens), then runs verify_visual.py, verify_docx.py, or verify_xlsx.py to catch clipped text, overlaps, and #### columns.
  • Codifies analysis discipline: treat anomalies as bugs first (Twyman's Law), prefer median/IQR, never use dual axes, plus seven "red lines" where everything looks fine while being wrong.

Who it's for

  • Anyone wrestling with system-exported spreadsheets full of merged cells and inconsistent formats.
  • Finance, sales, and marketing staff who must explain why two reports disagree on the same number.
  • Analysts and consultants who need executive-facing numbers to survive scrutiny.

Example uses

  1. "Accounting revenue is 3.2% off from CRM — find out why" → runs step 1, then step 5 reconciliation, checking denominator and refund-handling differences before assuming a calculation error.
  2. "Clean up this sales ledger" → repairs merged headers, thousands separators, and mixed date formats, and returns a replayable cleaning script with per-step affected row counts.
  3. "Build a monthly revenue report" → after aligning definitions, produces a self-contained HTML report with conclusion-first structure, inline SVG charts, and a passing visual verification run.

· · · Install guide · · ·

Try it now, no install

Paste this into Claude to use the skill without installing anything.

Read the instructions in this file and follow them to help me:
https://raw.githubusercontent.com/alchaincyf/huashu-excel/HEAD/SKILL.md

What I want: (describe your task here)

If Claude can't open the link, open it yourself and paste the contents instead.

↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.

Install in the Claude app (no terminal)
  1. Download the ZIP with the button below.
  2. In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
  3. Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
↓ Download ZIP
Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on AI Skill Mart.
Copy the SKILL.m folder from the GitHub repo alchaincyf/huashu-excel into my ~/.claude/skills/huashu-excel/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/alchaincyf/huashu-excel.git ~/.claude/skills/huashu-excel

⚠ This is a third-party skill. Check the source repository before installing.

  1. Open a terminal.
  2. Create the skills folder: mkdir -p ~/.claude/skills
  3. Clone the repo: git clone https://github.com/alchaincyf/huashu-excel.git ~/.claude/skills/huashu-excel
  4. Check your Python setup: python3 --version && python3 -c "import openpyxl, pandas". Install openpyxl (pip install openpyxl) if you need .xlsx reading/writing.
  5. Optional: for HTML/slide visual checks run pip install playwright && playwright install chromium. Without it verify_visual.py exits with code 2, meaning "nothing was verified".
  6. Restart Claude Code, then ask something like "profile and clean this Excel file first" to trigger the skill.
  7. For slides/PPT the skill hands off to huashu-design; install that repo separately when prompted.
View source on GitHub ↗License: MIT