The Agent Skills Directory

[PROMPT_INJECTION]: No attempts to override system instructions or safety filters were detected. The instructions maintain a consistent business diagnosis persona without trying to bypass agent constraints.
[DATA_EXFILTRATION]: No network operations, such as curl, wget, or fetch, are present. There are no attempts to access sensitive files like credentials or environment variables.
[COMMAND_EXECUTION]: The skill does not define any shell commands or subprocess executions. It operates entirely through natural language processing and dialogue management.
[REMOTE_CODE_EXECUTION]: No external scripts, remote code downloads, or third-party package installations are utilized.
[OBFUSCATION]: The content is provided in clear text with no evidence of Base64 encoding, zero-width characters, or hidden Unicode tags.
[INDIRECT_PROMPT_INJECTION]: While the skill ingests user-provided business scenarios, it lacks high-privilege tools (like file system writes or network access) that could be exploited. Furthermore, the skill's logic is explicitly designed to critically analyze and deconstruct user input for errors, which serves as a functional barrier against basic injection attempts.

dbs-diagnosis