Earnings Call Processor

Automated workflow: audio → transcript → financial data → structured analysis → Feishu draft.

Workflow

1. Transcribe audio — Use Whisper (or available transcription tool) to convert earnings call audio to text.
2. Fetch financial data — Use the finance skill tools (yfinance) to pull stock price, company info, and historical data for the target ticker.
3. Generate structured analysis — Combine transcript + financial data into a structured report. See references/analysis-template.md for the output structure.
4. Publish to Feishu — Use the feishu-doc skill to create a Feishu document draft with the analysis content.

Step-by-Step

1. Transcribe Audio

# If whisper is available
whisper <audio_file> --model turbo --language en --output_format txt --output_dir /tmp/whisper_out

# Otherwise, use any available transcription approach

If no audio file is provided, ask the user for the audio path or URL.

2. Fetch Financial Data

Use the finance skill's yfinance tools:

• yfinance_get_stock_price(symbol) — current price
• yfinance_get_stock_info(symbol) — company info, market cap, PE, etc.
• yfinance_get_historical_data(symbol, period="3mo") — recent price trend

3. Generate Structured Analysis

Read the transcript and financial data, then produce a structured analysis following the template in references/analysis-template.md.

Key sections:

• Executive Summary — 3-5 sentence overview
• Financial Highlights — revenue, EPS, guidance vs expectations
• Key Themes & Management Commentary — strategic priorities, risks, opportunities
• Q&A Highlights — notable analyst questions and management responses
• Stock Price Context — recent price action and valuation metrics
• Outlook & Takeaways — forward-looking assessment

4. Publish to Feishu

Use the feishu-doc skill:

python3 <feishu-doc-path>/scripts/doc_ctl.py create "财报电话会分析: {COMPANY} {QUARTER}" --content "<structured markdown>"

Return the Feishu document URL to the user.

Configuration

No special configuration required. Relies on:

• finance skill (for financial data)
• feishu-doc skill (for document publishing)
• Whisper or equivalent transcription tool (for audio)