The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructions focus on providing technical guidance for MySQL and do not contain patterns attempting to override agent behavior or safety guidelines.
[DATA_EXFILTRATION]: No evidence of unauthorized data transmission or access to sensitive local files for exfiltration purposes. The provided script templates use placeholder paths and domains for legitimate administrative tasks (e.g., backup alerting and object storage syncing).
[CREDENTIALS_UNSAFE]: No hardcoded sensitive credentials detected. Examples use clearly labeled placeholder values and document best practices for secret management, such as storing database passwords in separate files.
[REMOTE_CODE_EXECUTION]: No patterns of downloading and executing remote scripts or packages from untrusted sources. External tools referenced, such as Percona XtraBackup, Orchestrator, and gh-ost, are well-known and reputable services in the database community.
[COMMAND_EXECUTION]: The skill provides shell script templates for database administration tasks like backup, replication, and restoration. These are provided as informational examples for human implementation and are not configured for autonomous execution by the agent.
[EXTERNAL_DOWNLOADS]: References to external documentation and tools point to official repositories of well-known organizations and technology companies.

mysql