The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local Python script scripts/generate_config.py to generate tailored configuration files. This is the primary function of the skill and is handled via standard command-line arguments.
[PROMPT_INJECTION]: The configuration generation process is vulnerable to indirect prompt injection (Category 8). \n- Ingestion points: The generate_config.py script accepts various user-controlled parameters, such as cluster names, environment identifiers, and log file paths, which the agent provides based on user instructions. \n- Boundary markers: There are no boundary markers or delimiters used when interpolating these values into the INI-style configuration templates. \n- Capability inventory: The script writes a Fluent Bit configuration file (.conf) to the disk. This configuration defines log ingestion (file reads) and egress (network connections to cloud services and databases). \n- Sanitization: The script performs no validation or escaping on the input arguments; a malicious input could inject new configuration keys or sections into the resulting file.
[SAFE]: The skill correctly manages sensitive data by recommending environment variable placeholders (e.g., ${ES_USER}) instead of hardcoded secrets. While some examples use tls.verify Off, the documentation provides clear security guidance on using this setting only for internal or non-production environments.

fluentbit-generator