# What Are the Six Gates in the Investment Checklist Workflow? (AI-Berkshire Analysis) > Discover the six key gates in the AI-Berkshire investment checklist. Learn pass/fail criteria for Safety, Debt, Deterioration, Integrity, Insufficient-Data, and Margin-of-Safety gates for a robust analysis. - Repository: [Xbt Lin/ai-berkshire](https://github.com/xbtlin/ai-berkshire) - Tags: deep-dive - Published: 2026-07-26 --- **The AI-Berkshire repository implements a six-gate investment checklist where Safety, Debt, Deterioration, Integrity, and Insufficient-Data gates act as blocking filters that reject investments immediately upon failure, while the final Margin-of-Safety gate determines valuation-based acceptance.** The `xbtlin/ai-berkshire` open-source project codifies a defensive investment workflow that evaluates every potential opportunity against **six sequential risk gates**. This **investment checklist workflow** serves as a non-negotiable filter, ensuring that capital is only deployed when all fundamental safety, financial health, and data-quality criteria are satisfied. According to the repository's skill definitions, any failure in the first five gates triggers an immediate abort that overrides the quantitative scorecard. ## The Six Gates and Their Pass/Fail Criteria Each gate represents a distinct risk dimension defined in the project's skill files. The workflow evaluates these sequentially, with specific pass conditions and fail actions for each. ### 1. Safety Gate The **Safety Gate** tests for fundamental red flags related to regulatory compliance, environmental liabilities, or pending legal action. - **Pass Condition:** No safety-related red flags identified during due diligence. - **Fail Action:** Triggers an immediate **REJECT / YIELD TRAP** status, aborting the investment regardless of other metrics. ### 2. Debt Gate The **Debt Gate** evaluates capital structure sustainability by analyzing leverage ratios and coverage metrics. - **Pass Condition:** Debt metrics (e.g., debt-to-equity ratios, interest coverage) satisfy predefined thresholds indicating sustainable leverage. - **Fail Action:** Triggers **REJECT / YIELD TRAP**, preventing investment in over-leveraged entities. ### 3. Deterioration Gate The **Deterioration Gate** monitors for declining business fundamentals such as shrinking margins, falling cash flows, or eroding market position. - **Pass Condition:** Recent financial trends demonstrate stability or improvement across key operating metrics. - **Fail Action:** Triggers **REJECT / YIELD TRAP** when evidence of fundamental decline exists. ### 4. Integrity Gate The **Integrity Gate** screens for corporate governance risks, including fraud allegations, opaque ownership structures, or transparency deficiencies. - **Pass Condition:** No integrity concerns discovered during background and governance review. - **Fail Action:** Triggers **REJECT / YIELD TRAP** upon discovery of any material integrity issues. ### 5. Insufficient-Data Gate The **Insufficient-Data Gate** validates the availability and quality of primary source material required for informed decision-making. - **Pass Condition:** Evidence quality rated **A** (complete primary material) or **B** (partial primary material) according to the rating table defined in [`skills/income-investment.md`](https://github.com/xbtlin/ai-berkshire/blob/main/skills/income-investment.md) (lines 50-66). - **Fail Action:** Results in an **INSUFFICIENT DATA** outcome; the evaluation halts and cannot proceed until additional primary data is gathered. ### 6. Margin-of-Safety Gate The **Margin-of-Safety Gate** performs the final valuation assessment to determine if the price is sufficiently discounted relative to intrinsic value. This gate is described in [`README_EN.md`](https://github.com/xbtlin/ai-berkshire/blob/main/README_EN.md) (lines 449-452) as the ultimate valuation filter. - **Pass Condition:** Current price ≤ target valuation threshold (e.g., price-to-earnings ratios or discounted cash flow metrics). - **Fail Action:** Investment is marked **REJECT** (or **REDUCE** for existing positions) when the price lacks an adequate discount. ## Execution Flow and Gate Hierarchy The workflow distinguishes between **blocking gates** and the final valuation gate. According to the logic implemented in [`codex-skills/income-investment/SKILL.md`](https://github.com/xbtlin/ai-berkshire/blob/main/codex-skills/income-investment/SKILL.md) (lines 147-155), the first five gates (Safety, Debt, Deterioration, Integrity, and Insufficient-Data) function as mandatory filters. Any failure in these blocking gates **overrides** the quantitative scorecard and forces an immediate termination of the evaluation process. Only when **all five** blocking gates pass does the workflow proceed to the **Margin-of-Safety** assessment, which then determines whether the investment is accepted, reduced, or rejected based purely on valuation metrics. ## Evidence Quality Ratings The **Insufficient-Data Gate** relies on a specific grading system defined in [`skills/income-investment.md`](https://github.com/xbtlin/ai-berkshire/blob/main/skills/income-investment.md) to classify source material reliability: - **Grade A:** Complete primary material available (pass). - **Grade B:** Partial primary material available (pass). - **Grade C or lower:** Fail, triggering the insufficient data status. This rating system ensures that investment decisions are made only when supported by adequate factual basis. ## Summary - The **six gates** form a defensive perimeter: Safety, Debt, Deterioration, Integrity, Insufficient-Data, and Margin-of-Safety. - The first five gates operate as **blocking gates**; any single failure aborts the investment immediately according to [`codex-skills/income-investment/SKILL.md`](https://github.com/xbtlin/ai-berkshire/blob/main/codex-skills/income-investment/SKILL.md). - Evidence quality for the Insufficient-Data gate is strictly graded A or B as defined in [`skills/income-investment.md`](https://github.com/xbtlin/ai-berkshire/blob/main/skills/income-investment.md) (lines 50-66). - The Margin-of-Safety gate serves as the final valuation filter described in [`README_EN.md`](https://github.com/xbtlin/ai-berkshire/blob/main/README_EN.md) (lines 449-452). - Blocking gate failures override quantitative scores, ensuring qualitative risks take precedence over numerical metrics. ## Frequently Asked Questions ### What happens if an investment fails multiple blocking gates simultaneously? The workflow halts at the first failed gate encountered and returns the appropriate failure signal. Subsequent gates are not evaluated once a blocking condition is triggered, ensuring efficient rejection of unsuitable candidates. ### How does the Insufficient-Data gate differ from other blocking gates? While Safety, Debt, Deterioration, and Integrity gates trigger permanent **REJECT / YIELD TRAP** decisions, the Insufficient-Data gate triggers a temporary **INSUFFICIENT DATA** status. This allows the investment to be reconsidered when better primary material becomes available, rather than being permanently excluded. ### Where is the evidence quality rating system defined? The **Rate evidence quality** table specifying grades **A** and **B** resides in [`skills/income-investment.md`](https://github.com/xbtlin/ai-berkshire/blob/main/skills/income-investment.md) at lines 50-66. This table establishes the criteria for passing the Insufficient-Data gate. ### Can an investment pass all blocking gates but still be rejected? Yes. Passing the first five gates only permits progression to the final **Margin-of-Safety Gate**. If the current price exceeds the intrinsic value threshold defined in the valuation model, the investment receives a **REJECT** or **REDUCE** designation despite satisfying all safety and data-quality checks.