The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze and score prompts and test suites provided in text and JSON formats. This ingestion of external data creates a surface where malicious content within those files could attempt to influence the agent's interpretation of analysis results.
Ingestion points: The scripts eval_scorer.py, prompt_analyzer.py, and prompt_diff.py read content from files passed as command-line arguments.
Boundary markers: No explicit delimiters or boundary instructions are used when processing the content of these files.
Capability inventory: The skill is limited to reading local files and performing text analysis; it does not perform network operations or execute arbitrary code from these files.
Sanitization: No specific sanitization or escaping is performed on the ingested text before analysis.
[DYNAMIC_EXECUTION]: The eval_scorer.py utility dynamically compiles and executes regular expressions defined in the expected criteria of a user-provided JSON test suite. While a standard feature for pattern validation, it represents execution of logic defined in external data.
[COMMAND_EXECUTION]: The skill defines several CLI tools implemented as Python scripts. These are intended to be executed by the agent to perform prompt evaluation and analysis. The scripts are well-structured, use the standard library, and do not invoke dangerous shell subprocesses.

prompt-engineer-toolkit