config: add Keploy to LLM Tools collection

[TvisharajiK]

Keploy

Category: llm-dev-tools
Type: API Testing and Unit Testing Tool
Repository: keploy/keploy

Description:
Keploy is an open-source testing toolkit purpose-built for modern API and LLM-based applications. It captures real API calls and automatically generates test cases and mocks—making it ideal for validating LLM agents and pipelines. Keploy helps automate unit testing for LLM agents by recording their external calls (e.g., to language models, vector databases, or tools) and replaying them with mocks, enabling consistent and isolated testing without needing live dependencies.

Why it’s relevant to LLM Dev Tools:
Developers building applications with large language models (LLMs) often work with APIs, pipelines, and inference wrappers. Keploy helps in:

• Auto-generating unit and API tests
• Simulating API responses for local and CI testing

Key Features:

• 🧪 Auto-generates tests from real API traffic
• 🧱 Mocks upstream dependencies to isolate testing
• 🚀 Integrates seamlessly with CI/CD pipelines
• ⚙️ Supports any language or framework with HTTP interfaces

1. Run the agent once in a real or test environment—Keploy captures the outbound API requests (e.g., calls to OpenAI, tools, or databases).
2. Keploy automatically generates:
   • API tests that replicate full request-response interactions
   • Unit test mocks for each external dependency, isolating your agent logic
3. You can now run these tests locally or in CI without needing live API keys or internet access—making your LLM application reliable, reproducible, and testable end-to-end.