document-semantic-search

Search credit documents semantically to surface relevant clauses and covenants.

6|Updated Feb 5, 2026

Search Tags:#embeddings #knowledge-graph #semantic-search #collateral #covenants #document-preparation #credit-documents

Ready to run

npx skills add https://github.com/maschad/my-claude --skill document-semantic-search

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: document-semantic-search
Source: https://github.com/maschad/my-claude/tree/main/skills/document-semantic-search
Command: npx skills add https://github.com/maschad/my-claude --skill document-semantic-search

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Credit teams and deal lawyers spend excessive time locating relevant clauses, covenants, and terms across large credit-document corpora. This skill accelerates discovery by surfacing meaningfully similar provisions.

Core Features & Use Cases

Semantic search across credit documents (assignments, covenants, security, fee letters)
Metadata-aware filtering by company/deal
Covenant and collateral extraction for diligence workflows
Use case example: quickly locate leveraged ratio covenants across a deal pack and compare definitions

Quick Start

Query the corpus for 'financial covenants with leverage ratio' and review the top results.

Frequently Asked Questions about document-semantic-search

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

FAQPage Schema

How do I find specific covenants and terms across large credit document corpora?▼

To find specific covenants across credit document corpora, you can use semantic search to surface meaningfully similar provisions from assignments, credit agreements, and security documents. This accelerates discovery by locating relevant clauses based on meaning rather than exact keyword matches.

What is semantic search for credit documents and how does it work?▼

Semantic search for credit documents uses 4096-dim embeddings to identify and surface relevant clauses, covenants, and terms. It processes document chunking and metadata filters to deliver precise results with provenance tracking across financing materials.

Can I filter credit agreement clauses by specific company or deal metadata?▼

Yes, you can filter credit agreement clauses using metadata-aware filtering by company or deal. This allows you to narrow semantic search results to specific financing materials, supporting targeted covenant analysis and collateral assessment workflows.

What is the best way to compare financial covenant definitions like leverage ratios in a deal pack?▼

The best way to compare financial covenant definitions is querying the corpus for leverage ratio covenants and reviewing the top semantic results. This surfaces meaningfully similar provisions across the deal pack, enabling quick comparison of definitions and fee structures.

Does semantic search work for collateral assessment and fee structure benchmarking?▼

Yes, semantic search works for collateral assessment and fee structure benchmarking by extracting relevant covenants and terms from security documents and fee letters. It supports diligence workflows by surfacing precisely matched provisions across financing materials.

What types of credit documents are supported for semantic clause extraction?▼

Supported credit documents for semantic clause extraction include assignments, credit agreements, security documents, fee letters, and related financing materials. The skill applies 4096-dim embeddings to these documents to identify relevant provisions with provenance tracking.