The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied CSV files and incorporates the data into analytical reports, which creates a potential surface for indirect prompt injection if the input files contain malicious strings.
Ingestion points: The load_csv function in scripts/ts_analyzer.py reads data from user-specified file paths.
Boundary markers: Absent. The format_report function in scripts/ts_analyzer.py and the assets/ts_report_template.md do not use delimiters or warnings to isolate processed data from agent instructions.
Capability inventory: The skill has the capability to read local files (CSV) and write reports to the filesystem (via the --output flag in the Python script).
Sanitization: While metric values are validated and cast to floats, the date column values are processed as strings and included directly in the formatted output without escaping or sanitization.

time-series-analysis