13f analyzer

Skill cxcscmu/SkillLearnBench/skills/b4-skill-creator-claude-opus-4-6/financial-analysis/13f-analyzer

Perform data analysis on SEC 13-F filing datasets (TSV format) to obtain insights about fund activities such as number of holdings, AUM, and change of holdings between two quarters. Use this skill whenever analyzing hedge fund portfolios, comparing quarterly holdings, or working with SEC EDGAR 13F data files.From its SKILL.md

Install

npx -y skills add cxcscmu/SkillLearnBench --skill 13f-analyzer

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

SKILL.md

2.2 KB, 528 tokens by cl100k_base, as published. Nobody here has run it

13F Filing Analyzer

Dataset Structure

SEC 13F data comes as TSV files in quarterly folders (e.g., /root/2025-q2/, /root/2025-q3/):

File	Purpose	Key Columns
`COVERPAGE.tsv`	Filer metadata	`ACCESSION_NUMBER`, `FILINGMANAGER_NAME`, `REPORTCALENDARORQUARTER`
`INFOTABLE.tsv`	Individual holdings	`ACCESSION_NUMBER`, `NAMEOFISSUER`, `CUSIP`, `VALUE` (in thousands USD), `SSHPRNAMT` (shares)
`SUMMARYPAGE.tsv`	Filing summary	`ACCESSION_NUMBER`, `TABLEENTRYTOTAL` (count of holdings), `TABLEVALUETOTAL` (AUM in thousands)

Common Analysis Patterns

Get Fund's Accession Number

Search COVERPAGE.tsv by FILINGMANAGER_NAME to get the fund's ACCESSION_NUMBER for a given quarter.

Get AUM

Use the accession number to look up TABLEVALUETOTAL in SUMMARYPAGE.tsv. The value is in thousands of USD.

Get Number of Holdings

Use TABLEENTRYTOTAL from SUMMARYPAGE.tsv.

Compare Holdings Between Quarters

Get accession numbers for both quarters from COVERPAGE
Filter INFOTABLE by each accession number
Join on CUSIP and compute differences in VALUE or SSHPRNAMT
Rank by dollar value change (VALUE column, in thousands)

Find Top Holders of a Stock

Find the stock's CUSIP from INFOTABLE (search by NAMEOFISSUER)
Filter all INFOTABLE rows matching that CUSIP
Join with COVERPAGE on ACCESSION_NUMBER to get fund names
Rank by VALUE (in thousands)

Implementation Notes

Use Python with pandas for efficient TSV processing
TSV files use tab separator, read with pd.read_csv(path, sep='\t')
INFOTABLE is large (~3M rows); filter early to reduce memory usage
VALUE column is in thousands of USD
ACCESSION_NUMBER is the join key across all tables
When comparing quarters, some funds may have multiple filings (amendments); prefer non-amendment filings (ISAMENDMENT is empty or 'N')

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.