The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied RNA-seq expression matrices, group annotation files, and gene sets in scripts/io.R using read.csv and read.table.
Ingestion points: Data enters the system via the read_expression_matrix, read_group_info, and read_gene_sets functions.
Boundary markers: The instructions do not specify strict boundary markers for isolating the content of the processed CSV/TSV files.
Capability inventory: The R scripts possess capabilities to write files (CSV, RDS, PDF), create directories, and delete files within the specified output directory for cleanup purposes.
Sanitization: The implementation includes normalization of gene identifiers (casing) and sanitization of filenames to remove potentially problematic characters.
[EXTERNAL_DOWNLOADS]: The skill requires standard bioinformatics and data science packages from Bioconductor (GSVA) and CRAN (optparse, dplyr, tidyr, tibble, ggplot2, jsonlite, RColorBrewer, pheatmap). These are established and well-known software registries for the R ecosystem.
[EXTERNAL_DOWNLOADS]: The skill references the author's GitHub repository at github.com/aipoch/medical-research-skills as a resource for documentation and updates.
[COMMAND_EXECUTION]: The test script tests/run_tests.R executes the main entry point scripts/main.R using the system2 command to automate end-to-end validation. This is a standard practice for development and testing environments.

ssgsea-immune-infiltration-analysis