The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION] (INFO): The skill processes untrusted external data by reading the content of web files within a project directory.
Ingestion points: scripts/geo_checker.py reads file content using Path.read_text().
Boundary markers: None present; the script reads raw file content for analysis.
Capability inventory: The script is limited to local file reading and printing analysis results to stdout. It does not perform network operations, file writes, or command execution based on the file content.
Sanitization: Uses standard Python regex for pattern matching. No risk of execution from data.
[DATA_EXPOSURE] (LOW): The script reads local files to perform its audit. It includes a robust exclusion list (SKIP_DIRS, SKIP_FILES) to avoid sensitive directories like .git, node_modules, and configuration files, minimizing the risk of accidental exposure of secrets during the audit process.

geo-fundamentals