The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes shell commands and external binaries (such as pandoc and R) to process data and generate research artifacts. These operations are consistent with the skill's primary purpose of synthesizing medical data and producing PRISMA-compliant manuscripts.
[EXTERNAL_DOWNLOADS]: The workflow references standard external dependencies, including the python-docx Python library and specialized R statistical packages like mada, meta, metafor, and IPDfromKM. These are well-known tools in the medical research community and are used here to automate document formatting and statistical synthesis.
[PROMPT_INJECTION]: The skill ingests untrusted external data from literature searches and data extraction forms, which presents a surface for indirect prompt injection. To mitigate this risk, the instructions explicitly require independent human verification of all AI-generated content and establish multiple reconciliation gates to ensure numerical and methodological fidelity against primary source documents.

meta-analysis