agentleFS
zeiynzzeiynz/claude-skills/sec-filing-analyzer/SKILL.md

sec-filing-analyzer

Skill0 starsChanged 4 months agoMITView on GitHub

Extract, structure, and summarize SEC filings (10-K, 10-Q, 8-K, S-1) into research-ready formats — financial tables, risk factor deltas vs. prior filing, MD&A summaries, and insider-trading flags. Use this skill whenever the user mentions a ticker + filing type, pastes or links an SEC filing (EDGAR URL, accession number), asks to "summarize the 10-K", "what changed in the risk factors", "pull the financials from the latest 10-Q", or wants filing data turned into a structured research note. Also trigger for comparative filing analysis ("compare AAPL's last two 10-Ks") or building a filing-monitoring workflow.

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 4 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill for anything involving the structure of a filing (sections, XBRL tags, filing-to-filing deltas).

    line 13
  2. Distinctive:

    Only fall back to fetching the full filing HTML/text for qualitative sections (Risk Factors, MD&A, Legal Proceedings) that aren't in XBRL.

    line 31
  3. Distinctive:

    Default output is a research note with these sections (skip any that don't apply to the form type):

    line 34
  4. Distinctive:

    Default: structured markdown note (not a downloadable file) for quick research

    line 46
  5. Very specific:

    If user asks to "save" or wants a shareable doc: build an artifact or .docx using the docx skill, with the same section structure

    line 47
  6. Very specific:

    Fiscal year ≠ calendar year: check fiscalYearEnd before labeling periods

    line 51

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

SEC Filing Analyzer

Turns raw SEC filings into structured research artifacts: financial tables, risk-factor diffs, MD&A digests, and red-flag summaries. Built for retail/independent research workflows where a Bloomberg terminal isn't available.

When to use this vs. plain web search

Use this skill for anything involving the structure of a filing (sections, XBRL tags, filing-to-filing deltas). Don't use it for simple "what's the stock price" queries — that's a plain search.

Core workflow

Step 1: Resolve the filing

Given a ticker or company name, resolve to a CIK (Central Index Key) via EDGAR's company search:

https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&company={NAME}&type={FORM_TYPE}&dateb=&owner=include&count=10

Or if the user gives a direct EDGAR URL / accession number, fetch it directly. Always confirm which specific filing (company, form type, period, filing date) before deep analysis — ambiguity here silently produces wrong numbers.

Step 2: Fetch structured data first, HTML second

SEC's XBRL "company facts" API returns machine-readable financials — always prefer this over parsing the HTML filing for numbers:

https://data.sec.gov/api/xbrl/companyfacts/CIK{10-digit-cik}.json

This gives every reported XBRL fact (Revenues, NetIncomeLoss, Assets, etc.) with period and unit. Use it to build the financial summary table (see references/financial-tables.md for the standard table schema and common XBRL tag mappings).

Only fall back to fetching the full filing HTML/text for qualitative sections (Risk Factors, MD&A, Legal Proceedings) that aren't in XBRL.

Step 3: Structure the output

Default output is a research note with these sections (skip any that don't apply to the form type):

  1. Filing snapshot — company, CIK, form type, period, filed date, accession number
  2. Financial summary table — revenue, net income, EPS, cash, total debt, YoY deltas (see reference for exact line items by form type)
  3. Risk factor changes — if a prior filing is available, diff the risk factors section; call out new risks, removed risks, and materially reworded ones. Never reproduce the filing's risk factor text verbatim beyond a short paraphrase (copyright discipline applies even to public filings — SEC filings are public domain, so verbatim short quotes are fine here, but keep them minimal and cite the section).
  4. MD&A digest — 3-5 bullet summary of management's stated drivers, in your own words
  5. Flags — going concern language, restatements, auditor changes, insider 10b5-1 plan activity, unusual related-party transactions

Step 4: Comparative analysis (multi-filing requests)

For "compare last two 10-Ks" style requests, fetch both via the XBRL API in one pass (the companyfacts endpoint already returns multi-period history — you often don't need two separate fetches), then present as a comparison table rather than two separate notes.

Output format

  • Default: structured markdown note (not a downloadable file) for quick research
  • If user asks to "save" or wants a shareable doc: build an artifact or .docx using the docx skill, with the same section structure
  • Financial tables: always tabular, never prose-embedded numbers

Common pitfalls

  • Fiscal year ≠ calendar year: check fiscalYearEnd before labeling periods
  • Units matter: XBRL facts specify USD vs USD/shares — mislabeling this produces nonsense per-share figures
  • Amended filings (10-K/A) supersede the original — check for amendments before treating a filing as final
  • 8-Ks are event-driven, not periodic — don't apply the financial-summary-table workflow to an 8-K about an executive departure

Reference

  • references/financial-tables.md — standard table schemas per form type + common XBRL tag reference

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.