deep-company-series

Writes a 3-8 part long-form deep research article series analyzing a single public company.

16.4k|2.5k|Updated Apr 7, 2026
One-click install
npx skills add https://github.com/xbtlin/ai-berkshire --skill deep-company-series-xbtlin
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: deep-company-series
Source: https://github.com/xbtlin/ai-berkshire/tree/main/codex-skills/deep-company-series
Command: npx skills add https://github.com/xbtlin/ai-berkshire --skill deep-company-series-xbtlin

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Most AI-generated financial articles contain fabricated precision, subjective probability weighting, and absolute claims that cannot support real investment decisions. This Skill produces a rigorous multi-part company deep-dive series with enforced fact-checking, cross-article consistency, and value-investing discipline. ## Core Features & Use Cases - Adaptive Series Structure: Scales from 3 to 8 articles based on company complexity, covering moat analysis, profit engines, hidden assets, financials, management integrity, and valuation decisions. - Strict Fact-Check Checklist: Enforces a 7-point revision scan covering cross-article number consistency, accounting caliber labeling (Non-IFRS/GAAP/FCF), double-counting detection, and banned absolute wording. - Revision Workflow: Handles user feedback with a triage system (hard errors, subjective wording, granularity, unreliable third-party data) plus cascading updates when shared figures change. - Use Case: Ask for a deep series on a company like Tencent, and receive a structured set of publication-ready long-form articles saved under reports/{company}/ with a final privacy grep before pushing to GitHub. ## Quick Start Ask the AI to write a deep company series for a specific company, for example: write a deep-dive article series analyzing Tencent using the deep-company-series skill.

Frequently Asked Questions about deep-company-series

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I write a deep company research article series with AI?

Invoke this skill with a company name and it produces 3-8 long-form articles covering moat, profit engine, hidden assets, financials, management, and valuation. The article count adapts to company complexity, and each piece follows a fixed structural template with hooks and summaries.

How many articles should a company deep-dive series have?

The count depends on complexity: 7-8 articles for multi-business companies with large investment portfolios, 4-6 for mid-complexity firms, and 3 for focused single-business companies. Each article must stand alone around one sharp question.

What fact-checking rules apply to AI financial writing?

The skill enforces a 7-point checklist: cross-article number consistency, accounting caliber labeling, double-counting scans, fair peer comparisons, removal of probability-weighted expectations, elimination of absolute wording, and source attribution for third-party data.

When should I not use this deep company series skill?

Do not use it for single research reports, quarterly earnings reviews, or industry studies; those belong to dedicated research skills. It is designed only for multi-part public long-form series on one company.

How are user revision requests handled in financial articles?

Revisions are triaged into hard errors, subjective wording, granularity issues, and unreliable third-party data. Hard errors are fixed immediately, and any changed figure triggers cascading updates across all articles that reference it.