The Agent Skills Directory

[SAFE]: The skill documents internal infrastructure belonging to the vendor 'starchild-ai-agent' (sc-vpn.internal). No malicious patterns, obfuscation, or data exfiltration attempts were detected. All network operations are intended for bypassing region blocks within the platform's internal network environment.
[INDIRECT_PROMPT_INJECTION]: As the skill handles network routing for external requests via user-provided or agent-selected URLs, it possesses a standard indirect prompt injection attack surface.
Ingestion points: URLs and request targets passed to the requests library or curl command.
Boundary markers: None explicitly defined in the skill documentation.
Capability inventory: Outbound HTTP/HTTPS network requests using curl and requests.get().
Sanitization: Not applicable at the skill definition level; relies on platform-level network security and agent guardrails.

sc-vpn