メニュー
Extract business rules, boundary conditions, and exception handling strategies from source code across any language. Produces structured, auditable documentation for knowledge preservation, migration planning, and team onboarding.
デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。
インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
直接コマンドでは確認用 Prompt が省略されます。実行前にソースを確認してください。
npx skills add https://github.com/pan94u/forge --skill business-rule-extractionコマンドは1行のまま表示されます。コピー前に横へスクロールして全体を確認してください。
ローカルで確認しますか？SkillsMP が現在取得できるファイルをダウンロードできます。
Planning Mode skill for guided execution of large tasks. Provides structured task decomposition, user confirmation, per-task verification, retry logic, and quality summary.
Git 工作流 — 版本控制、提交规范、分支管理、.gitignore 治理
Structured bug diagnosis and repair workflow using the 5-Whys root cause analysis method. Ensures fixes include tests, documentation, and knowledge capture.
SOC 職業分類に基づく
SKILL.md を表示中
| name | business-rule-extraction |
| description | Extract business rules, boundary conditions, and exception handling strategies from source code across any language. Produces structured, auditable documentation for knowledge preservation, migration planning, and team onboarding. |
| trigger | when the user requests business rule extraction, domain logic analysis, or migration knowledge mining |
| tags | ["business-rules","domain-logic","knowledge-mining","migration","analysis"] |
| profile | planning |
| supported-languages | ["java","kotlin","csharp","python","go","typescript"] |
| version | 2.0 |
| scope | platform |
| category | foundation |
Systematically extract and document business rules embedded in source code. This is critical for:
Scan for classes/functions that contain domain logic (not infrastructure):
| Language | Look For |
|---|---|
| Java/Kotlin | *Service.java, *Validator.java, *Rule*.java, *Policy*.java, *Calculator*.java |
| C# | *Service.cs, *Validator.cs, *Handler.cs (MediatR), *Specification.cs |
| Python | services/*.py, rules/*.py, validators/*.py, domain/*.py |
| Go | *_service.go, *_handler.go, *_validator.go |
Skip infrastructure code: controllers, repositories, DTOs, config, migrations.
For each business method, extract:
Conditions and logic that enforce business policies.
Rule: BR-ORDER-001
Source: OrderService.kt:45 — calculateTotal()
Description: Orders over $500 qualify for free shipping
Condition: order.subtotal > 500.00
Action: shipping cost set to 0
Edge cases and limits the system enforces.
Boundary: BC-ORDER-001
Source: OrderService.kt:52 — validateOrder()
Description: Maximum 99 items per order
Condition: order.items.size > 99
Action: throw MaxItemsExceededException
How the system responds to failures.
Error Strategy: ES-PAYMENT-001
Source: PaymentService.kt:78 — processPayment()
Trigger: Payment gateway returns HTTP 503
Strategy: Retry 3 times with exponential backoff, then mark order as "payment_pending"
Compensation: Scheduled job retries after 15 minutes
Organize extracted rules by bounded context or business domain:
orders/
business-rules.md — All order-related rules
boundary-conditions.md — All order limits and edge cases
error-strategies.md — All order error handling
payments/
business-rules.md
boundary-conditions.md
error-strategies.md
inventory/
...
# {Domain} Business Rules
## Summary
- Total rules extracted: {N}
- Total boundary conditions: {N}
- Total error strategies: {N}
- Test coverage: {covered}/{total} rules have corresponding tests
## Business Rules
### BR-{DOMAIN}-001: {Short Description}
- **Source**: `{file}:{line}` — `{methodSignature}`
- **Description**: {Business-language description of the rule}
- **Condition**: `{code condition or pseudocode}`
- **Action**: {What happens when condition is true}
- **Test**: {link to test or "UNTESTED"}
- **Confidence**: HIGH | MEDIUM | LOW
- **Notes**: {Any ambiguity or assumptions}
### BR-{DOMAIN}-002: ...
## Boundary Conditions
### BC-{DOMAIN}-001: {Short Description}
- **Source**: `{file}:{line}` — `{methodSignature}`
- **Type**: MIN_VALUE | MAX_VALUE | REQUIRED | FORMAT | STATE_PREREQUISITE
- **Constraint**: `{the actual constraint}`
- **Violation Response**: {exception thrown or error returned}
- **Test**: {link to test or "UNTESTED"}
## Error Handling Strategies
### ES-{DOMAIN}-001: {Short Description}
- **Source**: `{file}:{line}` — `{methodSignature}`
- **Trigger**: {What error condition triggers this strategy}
- **Strategy**: RETRY | COMPENSATE | FAIL_FAST | DEGRADE | ESCALATE
- **Details**: {Retry count, backoff policy, fallback behavior, etc.}
- **Compensation**: {Rollback or compensation action, if any}
- **Alert**: {Whether ops is notified}
# Business Rule Extraction Summary
## Extraction Metadata
- System: {system name}
- Date: {extraction date}
- Extractor: Forge AI + {human reviewer}
- Source languages: {languages found}
- Modules analyzed: {count}
## Rule Statistics
| Domain | Business Rules | Boundaries | Error Strategies | Test Coverage |
|--------|---------------|------------|-----------------|---------------|
| Orders | 15 | 8 | 5 | 73% |
| Payments | 12 | 6 | 8 | 80% |
| Inventory | 9 | 4 | 3 | 56% |
| **Total** | **36** | **18** | **16** | **70%** |
## Cross-Domain Dependencies
- Order creation triggers inventory reservation (BR-ORDER-003 → BR-INV-001)
- Payment failure triggers order cancellation (ES-PAY-002 → BR-ORDER-010)
## Risk Areas
- {Rules with LOW confidence — need human validation}
- {Rules with no tests — high migration risk}
- {Conflicting rules across domains}
Before the extraction is considered complete, a human domain expert MUST review:
| Criterion | Threshold |
|---|---|
| Coverage | 100% of Service/Handler classes analyzed |
| Completeness | Every extracted rule has: source, description, condition, action |
| Traceability | Every rule links to exact file:line |
| Reviewability | Descriptions use business language, not code jargon |
| Confidence | No more than 20% of rules rated LOW confidence |