The Agent Skills Directory

[COMMAND_EXECUTION] (HIGH): The skill enables the agent to perform credential brute-forcing and authentication testing against local or remote ONVIF devices using the onvifscan tool. Providing an AI agent with automated attack capabilities on network infrastructure is considered a high-risk capability.
[EXTERNAL_DOWNLOADS] (MEDIUM): The skill relies on an external binary located in the iothackbot directory. This is an unverifiable dependency whose source and integrity are not established within the skill's context, posing a risk of executing unverified code.
[PROMPT_INJECTION] (LOW): Indirect Prompt Injection Surface detected. The tool ingests and potentially displays XML responses from untrusted network devices, which could be used to influence agent behavior.
Ingestion points: XML data returned from ONVIF device endpoints during scanning (onvifscan auth/brute).
Boundary markers: Absent. The skill does not instruct the agent to treat device output as untrusted or provide delimiters for external data.
Capability inventory: Network scanning, credential brute-forcing, and command execution via the onvifscan binary.
Sanitization: Absent. No evidence of output validation or escaping before the LLM processes device responses.

onvifscan