The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill generates executable scripts (.R, .py, or .do files) at runtime to perform statistical calculations. This dynamic script generation is based on templates that incorporate user-provided variables such as outcome names, estimands, and design parameters.
[INDIRECT_PROMPT_INJECTION]: The skill defines an attack surface where untrusted data is ingested into the agent context and used in high-capability operations.
Ingestion points: Data is read from user-supplied arguments and external specification files (e.g., quality_reports/specs/) generated by other tools.
Boundary markers: The workflow does not specify delimiters or instructions to prevent the agent from obeying command-like content embedded within these data sources.
Capability inventory: The skill has access to the Write tool to create scripts and the Bash and Task tools to execute them on the local system.
Sanitization: There is no mention of escaping, filtering, or validating the input data before it is interpolated into the generated scripts.
[COMMAND_EXECUTION]: The skill uses shell-level tools (Bash, Task) to execute generated scripts and run simulation harnesses, which is a necessary but high-privilege operation for its primary purpose.

power-analysis