Build AI pipelines with prompts that improve.
Guides the construction and data-driven improvement of modular language-model pipelines with DSPy.
When to use it
Use when building multi-stage AI systems, replacing manual prompt engineering, or improving model outputs through optimization.
It changes how the agent builds language-model pipelines, starting with simple predictions and adding reasoning and data-driven optimization as needed.
What you provide
The module examples use Python classes and control flow; no runtime version is specified.
The guide's language-model programs import and use the dspy package.
The Claude configuration used in the quick start accepts ANTHROPIC_API_KEY from the environment as an alternative to supplying the api_key argument.
This skill
dspy
Installs the dspy package