The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill communicates with Volcengine's official API endpoint at open.volcengineapi.com to retrieve monitoring metrics and configuration status. These network operations are legitimate requests to the vendor's own infrastructure.
[INDIRECT_PROMPT_INJECTION]: The skill processes data returned from external AntiDDoSPro API responses.
Ingestion points: Data is fetched from the network via the diagnostic functions defined in scripts/advdefence_client.py.
Boundary markers: The SKILL.md file provides explicit instructions for the agent to be conservative in its analysis, requiring it to distinguish between 'no data' and 'no risk'.
Capability inventory: The skill utilizes the requests library to perform authenticated GET and POST operations against vendor infrastructure.
Sanitization: The agent is instructed to summarize the findings from the parsed JSON data rather than outputting raw payloads.
[SAFE]: The skill implements secure credential management by fetching required API keys (VOLC_ACCESS_KEY and VOLC_SECRET_KEY) from system environment variables, avoiding the use of hardcoded secrets or sensitive inputs in the conversation context.

byted-antiddospro-hostops