Compile declarative DSPy programs into optimizable pipelines from labeled examples.
npx skills add https://github.com/devMoez/titan --skill dspy-devmoez
Please help me install this Agent Skill. Skill: dspy Source: https://github.com/devMoez/titan/tree/main/skills/mlops/research/dspy Command: npx skills add https://github.com/devMoez/titan --skill dspy-devmoez
DSPy removes brittle, manual prompt engineering by letting you define LM programs declaratively and automatically optimize them using feedback from data.
Use Case: You want a higher-accuracy RAG QA system that retrieves relevant passages, generates grounded answers, and improves its prompting using a small set of question/answer examples.
Use DSPy to build a RAG question-answering pipeline that defines a signature, configures a retriever, and then compiles the module using BootstrapFewShot with your custom exact-match metric.
High-intent search queries and answers about installing and using this skill.
You can auto-optimize LLM pipelines by defining declarative signatures and modules, then using teleprompters to compile them against labeled examples and evaluation metrics. This replaces manual prompt engineering with automated instruction and few-shot demo improvement.
Build a reliable RAG QA system by defining a DSPy signature for question answering, configuring a retriever for relevant passages, and compiling the module using an optimizer like BootstrapFewShot with a custom exact-match metric to improve grounded answer generation.
Teleprompting is the process of automatically optimizing LM pipeline instructions and few-shot demonstrations using an optimizer and evaluation metric. It compiles declarative modules against labeled examples to improve performance for tasks like classification, extraction, and multi-stage reasoning.
Yes, you need labeled examples and an evaluation metric to compile LM pipelines. The optimizer uses these examples to automatically improve instructions and few-shot demos, ensuring portable and maintainable prompts for RAG, classification, and agent workflows.
Yes, DSPy works with OpenAI and Anthropic models. These dependencies allow you to declaratively build and auto-optimize language model pipelines across different providers for classification, extraction, and multi-stage reasoning tasks.
The best way to maintain prompt portability is to define declarative signatures and composable modules for your multi-stage reasoning tasks. Compiling these modules with an optimizer ensures prompts are automatically improved and remain maintainable across different model calls.
8b-isBuild type-safe DSPy.rb workflows with signatures, modules, and provider configurations.
zerwizBuild declarative LLM programs and RAG pipelines with DSPy.
cabeznoDefine declarative LM programs and optimize prompts with DSPy.
tiandiyiqiOptimize LLM prompts and design agentic AI systems.
