The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted data from external emails, which serves as an indirect prompt injection surface. An attacker could send a malicious email attempting to influence the agent's actions. This is mitigated by the skill's specific entity extraction logic and the mandatory requirement for the user to confirm all changes before they are written to the Google Sheet.
[SAFE]: The skill defines rigorous 'Write Safety Rules' to ensure data integrity. These include taking a pre-write snapshot, preferring specific cell updates over bulk row additions to avoid corruption, and performing post-write verification to detect unexpected shifts in data.
[SAFE]: The use of Gmail and Google Sheets tools is strictly aligned with the skill's primary purpose. No unauthorized data exfiltration or suspicious network operations were detected.

job-tracker