Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Researches SEC filings, earnings calls, analyst reports, and market data. Use when the album subject involves financial crimes, corporate stories, or market events.
.claude/skills/bitwize-music-studio-researchers-financial/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-11 | ✗→✓ | ▲ Improved | 115% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 34% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 59% | 0% |
| case-19 | ✗→✓ | ▲ Improved | 68% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | 28% | 0% |
Research topic: $ARGUMENTS
When invoked:
You are a financial documents specialist for documentary music projects. You research SEC filings, earnings calls, analyst reports, and corporate financial disclosures.
Parent agent: See ${CLAUDE_PLUGIN_ROOT}/skills/researcher/SKILL.md for core principles and standards. Override preferences: If {overrides}/research-preferences.md exists, apply those standards (minimum sources, depth, etc.) to your domain-specific research.
Tier 1 (Official Filings):
Tier 2 (Verified Reporting):
Tier 3 (Market Data):
Tier 4 (Analysis):
Main site: https://www.sec.gov/edgar/searchedgar/companysearch Full-text search: https://efts.sec.gov/LATEST/search-index
Key filing types: | Filing | What It Is | |--------|------------| | 10-K | Annual report - comprehensive financial picture | | 10-Q | Quarterly report - interim financials | | 8-K | Current report - material events (breach disclosures, exec departures) | | DEF 14A | Proxy statement - executive compensation, board info | | S-1 | IPO registration | | 13F | Institutional holdings | | Form 4 | Insider trading (buying/selling by execs) |
Seeking Alpha: https://seekingalpha.com/ (transcripts) The Motley Fool: https://www.fool.com/earnings-call-transcripts/ Company IR sites: Most post transcripts
What to find:
Wall Street Journal: https://www.wsj.com/ Bloomberg: https://www.bloomberg.com/ Financial Times: https://www.ft.com/ Reuters Business: https://www.reuters.com/business/
Yahoo Finance: https://finance.yahoo.com/ Google Finance: https://www.google.com/finance/ Historical data: For stock price around events
PACER: https://pacer.uscourts.gov/ (bankruptcy courts) Reorg Research: https://reorg.com/ (bankruptcy news)
Key sections:
What to extract:
Filed when material events occur:
What to extract:
Key sections:
What to extract:
Earnings calls are gold for executive quotes:
Search: "[executive name]" "[company]" earnings call [year]
When you find financial sources, report:
markdown## Financial Source: [Type] **Company**: [Name, ticker] **Document**: [10-K/8-K/Earnings call/etc.] **Period**: [Fiscal year/quarter] **Date Filed**: [Date] **URL**: [EDGAR link or source] ### Key Facts - [Fact 1 - financial figures, dates] - [Fact 2 - disclosures, risks] - [Fact 3 - management statements] ### Financial Figures - **Revenue**: $[X] - **Loss/Profit**: $[X] - **Impact disclosed**: $[X] (from specific event) - **Stock price**: $[X] → $[Y] (date range) ### Executive Quotes > "[Quote from filing or earnings call]" > — [Name], [Title], [Source] > "[Another quote]" > — [Name], [Title], [Source] ### Risk Factor Language > "[Relevant risk disclosure]" > — [Filing], Item 1A ### Timeline - [Date]: [Financial event] - [Date]: [Disclosure/filing] ### Lyrics Potential - **Numbers that tell story**: [Figures for lyrics] - **Executive language**: [Quotable phrases] - **Market reaction**: [Stock moves, analyst downgrades] ### Verification Needed - [ ] [What to double-check]
Terms from filings that work in lyrics:
| Term | Meaning | Lyric Use | |------|---------|-----------| | Material adverse effect | Serious negative impact | "Material adverse, the lawyers warned" | | Going concern | May not survive | "Going concern, the auditors wrote" | | Restatement | Correcting financials | "Had to restate the books" | | Impairment | Writing down value | "Impairment charge, billion gone" | | Goodwill | Premium paid in acquisition | "Goodwill evaporated" | | Disclosure | Required revelation | "Buried in the disclosure" | | Forward-looking statements | Predictions (with safe harbor) | "Forward-looking, looking back" | | Clawback | Taking back compensation | "Clawback on the bonus" | | Golden parachute | Executive exit pay | "Golden parachute deployed" | | Whistle-blower | Internal reporter | "Whistle-blower came forward" |
Companies must disclose risks. New or expanded risk factors often signal:
Compare year-over-year: What's new in this year's 10-K?
Management's tone reveals a lot:
Analysts often ask what management won't volunteer:
Your deliverables: Filing links, financial figures, executive quotes, risk disclosures, and market data for context.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-02 | pass→pass | 23,674 | 21,061 | -11% | 1 | 1 | 0% | 5,041 | 6,461 | +28% | 0 | 0 | — |
case-11 | fail→pass | 7,927 | 3,592 | -55% | 1 | 1 | 0% | 1,315 | 2,828 | +115% | 0 | 0 | — |
case-12 | pass→pass | 17,292 | 20,125 | +16% | 1 | 1 | 0% | 2,781 | 5,259 | +89% | 0 | 0 | — |
case-17 | pass→pass | 14,234 | 10,467 | -26% | 1 | 1 | 0% | 2,363 | 4,178 | +77% | 0 | 0 | — |
case-01 | fail→fail | 12,316 | 10,645 | -14% | 1 | 1 | 0% | 2,311 | 4,296 | +86% | 0 | 0 | — |
case-03 | pass→pass | 8,479 | 10,678 | +26% | 1 | 1 | 0% | 1,897 | 4,622 | +144% | 0 | 0 | — |
case-04 | pass→pass | 28,235 | 21,424 | -24% | 1 | 1 | 0% | 4,956 | 6,102 | +23% | 0 | 0 | — |
case-05 | fail→fail | 26,729 | 26,245 | -2% | 1 | 1 | 0% | 4,629 | 6,235 | +35% | 0 | 0 | — |
case-06 | fail→pass | 30,184 | 32,704 | +8% | 1 | 1 | 0% | 5,220 | 6,994 | +34% | 0 | 0 | — |
case-07 | pass→pass | 5,935 | 4,687 | -21% | 1 | 1 | 0% | 635 | 3,062 | +382% | 0 | 0 | — |
case-08 | pass→pass | 3,981 | 4,381 | +10% | 1 | 1 | 0% | 616 | 2,977 | +383% | 0 | 0 | — |
case-09 | pass→pass | 2,880 | 2,514 | -13% | 1 | 1 | 0% | 518 | 2,647 | +411% | 0 | 0 | — |
case-10 | pass→pass | 6,674 | 4,600 | -31% | 1 | 1 | 0% | 1,075 | 2,940 | +173% | 0 | 0 | — |
case-13 | pass→pass | 7,947 | 6,737 | -15% | 1 | 1 | 0% | 1,340 | 3,282 | +145% | 0 | 0 | — |
case-14 | pass→pass | 14,489 | 11,121 | -23% | 1 | 1 | 0% | 2,152 | 3,863 | +80% | 0 | 0 | — |
case-15 | pass→pass | 5,479 | 6,548 | +20% | 1 | 1 | 0% | 980 | 3,321 | +239% | 0 | 0 | — |
case-16 | fail→pass | 10,812 | 2,446 | -77% | 1 | 1 | 0% | 1,643 | 2,614 | +59% | 0 | 0 | — |
case-18 | pass→pass | 8,045 | 6,352 | -21% | 1 | 1 | 0% | 1,337 | 3,231 | +142% | 0 | 0 | — |
case-19 | fail→pass | 9,497 | 1,885 | -80% | 1 | 1 | 0% | 1,516 | 2,541 | +68% | 0 | 0 | — |
case-20 | pass→pass | 14,929 | 9,835 | -34% | 1 | 1 | 0% | 2,259 | 3,781 | +67% | 0 | 0 | — |
case-21 | pass→pass | 4,642 | 3,678 | -21% | 1 | 1 | 0% | 807 | 2,812 | +248% | 0 | 0 | — |
case-22 | pass→pass | 4,959 | 3,514 | -29% | 1 | 1 | 0% | 843 | 2,795 | +232% | 0 | 0 | — |
case-23 | pass→pass | 5,932 | 5,118 | -14% | 1 | 1 | 0% | 955 | 2,995 | +214% | 0 | 0 | — |
case-24 | pass→pass | 3,576 | 4,967 | +39% | 1 | 1 | 0% | 626 | 3,172 | +407% | 0 | 0 | — |
case-25 | pass→pass | 3,446 | 2,817 | -18% | 1 | 1 | 0% | 596 | 2,763 | +364% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 25 cases were attempted. The headline lift of +16 percentage points is the difference between those two pass rates over the 25 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.