The Agent Skills Directory

[PRIVILEGE_ESCALATION]: The skill's infrastructure definitions request elevated Linux capabilities for containerized network operations.
Evidence: cap_add: [NET_ADMIN] in the proxy_tunnel service definition within SKILL.md used for Wireguard client operations.
[EXTERNAL_DOWNLOADS]: The skill instructs the agent to download and use external Docker images and interact with remote API services.
Evidence: image: camoufox/stealth:latest and image: wireguard/client:latest in SKILL.md.
Evidence: External API call to api.ipdata.co in the probe_proxy_health Python snippet.
[INDIRECT_PROMPT_INJECTION]: The skill provides a health probe mechanism that ingests JSON data from an external IP intelligence API. This data represents an attack surface where malicious or unexpected content could influence downstream agent logic if not properly sanitized.
Ingestion points: probe_proxy_health function in SKILL.md (ingests response from api.ipdata.co).
Boundary markers: Absent in the provided Python snippet code.
Capability inventory: execute_command, write_file, edit_file, create_file as defined in the YAML frontmatter.
Sanitization: Absent in the snippet; the skill relies on the ASI05 RCE Guard textual instruction for manual verification.

deploy-mmo-infrastructure