The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the Bash tool to execute a local Python script, scripts/gdpr_deadlines.py, for calculating GDPR-compliant deadlines.
The execution is limited to a specific script provided within the skill package.
The script uses standard libraries and performs deterministic date arithmetic without external network or file system dependencies.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process external user data in the form of DSAR requests, which represents a potential injection surface.
Ingestion points: Untrusted data enters the context via the content of the data subject's request (e.g., "wniosek o dostęp do danych").
Boundary markers: The skill relies on the agent's internal reasoning and structured instructions rather than explicit delimiters like XML tags or triple quotes to isolate user input.
Capability inventory: The skill has access to the Bash tool to run a local deadline calculator and the Read tool to access project files.
Sanitization: The deadline calculator script employs strict date parsing using datetime.strptime, which prevents arbitrary command injection through the date arguments passed to the script.
[SAFE]: The skill follows security best practices by setting requires-human-approval: true and explicitly leaving irreversible actions (deletion, export, sending emails) to a human operator. It hardcodes no secrets and makes no network connections.

rodo-dsar-pl