The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill provides documentation for the dspy.ProgramOfThought module, which is designed to generate and execute Python code at runtime for solving computational or logical problems.
[DYNAMIC_EXECUTION]: Educational examples in references/examples.md demonstrate a custom calculate tool that uses the eval() function to evaluate string-based mathematical expressions.
[INDIRECT_PROMPT_INJECTION]: The skill documents architectural patterns for Retrieval-Augmented Generation (RAG) and search-based agents that process data from external, untrusted sources such as Wikipedia or web search results.
Ingestion points: External content is ingested via dspy.Retrieve modules and search tool integrations described in SKILL.md and references/examples.md.
Boundary markers: The provided code snippets do not include explicit prompt delimiters or system instructions to mitigate the risk of the model following instructions embedded within the retrieved data.
Capability inventory: The documented patterns include high-impact capabilities such as dynamic code execution and external network calls.
Sanitization: The instructions do not demonstrate explicit sanitization, filtering, or validation of the retrieved content before it is processed by the language model.
[EXTERNAL_DOWNLOADS]: The skill includes instructions to install the dspy library directly from the official Stanford NLP GitHub repository and other standard package registries.

dspy