agentleFS
efaslantasefaslantas/claude-finops/.claude/skills/pdf-analyzer/SKILL.md

pdf-analyzer

Skill2 starsChanged 5 months agoMITView on GitHub

Read a financial PDF (annual report, 10-K/10-Q, prospectus, bank statement) using Claude's native long-context + vision capabilities. Extracts structured financials, key risks, management commentary and red flags. No OCR tool needed — Claude reads the document directly.

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read the file with the Read tool (PDF is natively supported).

    line 27

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

PDF Financial Analyzer

What Claude does here that other tools cannot: Claude's 200 000-token context window reads an entire annual report in one pass — not just a summary or the first 10 pages. Vision capability handles scanned tables, charts and mixed-format filings without pre-processing.

Trigger phrases

  • "Bu PDF'i analiz et: <dosya-yolu>"
  • "Analyze this annual report: <path>"
  • "10-K/yıllık raporu oku ve özetle"
  • "Banka ekstresini parse et"

Phases

Phase 1 — Load & classify

Read the file with the Read tool (PDF is natively supported). Determine document type: annual_report | 10k | 10q | prospectus | bank_statement | other. Note language, currency, fiscal year end.

Phase 2 — Extract financials (structured)

For annual reports / 10-K / 10-Q, extract into this JSON structure:

{
  "company": "",
  "period": "",
  "currency": "",
  "income_statement": {
    "revenue": null,
    "gross_profit": null,
    "ebitda": null,
    "ebit": null,
    "net_income": null,
    "eps": null
  },
  "balance_sheet": {
    "total_assets": null,
    "total_debt": null,
    "cash": null,
    "equity": null,
    "net_debt": null
  },
  "cash_flow": {
    "operating_cf": null,
    "capex": null,
    "free_cash_flow": null,
    "dividends_paid": null
  },
  "ratios": {
    "pe": null,
    "ev_ebitda": null,
    "debt_equity": null,
    "current_ratio": null,
    "roe": null,
    "net_margin_pct": null
  }
}

For bank statements, extract: {period, opening_balance, closing_balance, total_in, total_out, transactions[]}.

Phase 3 — Key risks & management commentary

  • Identify top 5 risks stated in the document (section: Risk Factors / Riskler)
  • Summarize CEO/management outlook in 3 bullet points
  • Flag any going-concern, restatement, regulatory or litigation disclosures

Phase 4 — Red flags

Cross-check for:

  • Revenue recognition policy changes (IFRS 15 / ASC 606)
  • Debt covenant disclosures
  • Qualified / adverse auditor opinion
  • Unusual related-party transactions
  • Free cash flow ≠ net income divergence > 30%
  • Goodwill impairment risk (goodwill > 40% of total assets)

Phase 5 — Output

Write output/pdf-<company>-<YYYYMMDD>.md with:

# PDF Analysis: <Company> (<Period>)

## Income Statement (summary)
## Balance Sheet (summary)
## Cash Flow (summary)
## Key Ratios
## Top 5 Risks
## Management Commentary
## Red Flags
## Raw Extracted JSON

Assumptions & limitations: OCR quality varies on scanned PDFs. All figures as-reported; no currency conversion unless base_currency is specified. Mark any extracted value with * if confidence is low (table layout complex or partially scanned).

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.