The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from a local SQLite database (puda.db), which serves as a vulnerability surface for indirect prompt injection.
Ingestion points: Data is ingested from the payload column of the command_log table as described in references/analysis.md.
Boundary markers: The instructions do not specify the use of delimiters or specific warnings to ignore instructions embedded within the database data.
Capability inventory: The skill possesses capabilities to execute shell commands (puda CLI), install external packages via pip, and write/execute Python scripts.
Sanitization: There is no evidence of sanitization or filtering applied to the database content before it is parsed as JSON or processed by pandas.
[DYNAMIC_EXECUTION]: The skill workflow involves generating Python scripts from templates to perform data analysis and plotting, which are then executed within an analysis/ directory.
[COMMAND_EXECUTION]: The skill utilizes the puda command-line utility for database schema inspection, project data extraction, and machine command lookups.

puda-report