Stock Qualitative Analysis

What it does

Provides a structured, chapter-by-chapter qualitative analysis workflow for public companies. The skill ingests filings (PDF/HTML) and public sources, extracts key sections (e.g., Item 1/1A/7/8), and produces a Markdown report with Conclusions / Details / Evidence for each section. It enforces citation-first writing and placeholders when evidence is missing, and defaults to Chinese output unless English is requested.

When to use it

Use this skill when a user requests an evidence-backed qualitative stock analysis (定性分析) that must avoid hallucinations and follow a buy-side reporting format — e.g., research notes, company deep-dives, or due-diligence summaries. Suitable when SEC filings or user-provided PDFs are available.

What's included

• Scripts: mentions of scripts to fetch and preprocess sources (e.g., build_source_manifest.py, extract_sec_html_sections.py) though repository has no scripts flagged by fetch.
• References: report template in assets/report-template.md and guardrails for citation formatting.
• Instructions: procedural intake checklist, pre-check local data, acquire sources, section-by-section generation, progressive write-back, and finalization steps.

Compatible agents

Intent: tools that can run Python scripts and fetch remote filings (agents with web/network access, Claude/Codex-style agents that can call helper scripts).