The Agent Skills Directory

[PROMPT_INJECTION]: The skill ingests network state data (such as BGP states, interface descriptions, and hostnames) from external devices. This creates an indirect prompt injection surface where an attacker could influence the agent by placing malicious instructions in device configuration fields. \n
Ingestion points: Tool outputs from suzieq_show, suzieq_summarize, suzieq_unique, and suzieq_path that return device-level data.\n
Boundary markers: None specified in the skill instructions to delimit untrusted network data or warn the agent against executing instructions contained within.\n
Capability inventory: The skill provides extensive read access to over 20 network tables, allowing for wide exposure to potentially malicious configuration data.\n
Sanitization: No data validation or sanitization mechanisms are defined to filter external network content before it enters the agent's context.\n- [DATA_EXFILTRATION]: The skill uses environment variables (SUZIEQ_API_URL, SUZIEQ_API_KEY) for authentication with the SuzieQ API. This is a recommended security practice for handling sensitive credentials and avoids hardcoded secrets. The skill performs network operations to the user-provided API URL to retrieve observability data.

suzieq-observability