The Agent Skills Directory

[SAFE]: The skill consists of technical documentation and configuration examples. It contains no executable scripts, binaries, or hidden payloads that would pose a risk to the environment.
[DATA_EXFILTRATION]: Configuration examples for HTTP proxies and Redis connections correctly reference environment variables (e.g., ENV["HTTP_PROXY"], ENV["REDIS_URL"]) rather than hardcoding credentials, adhering to standard security best practices.
[PROMPT_INJECTION]: The documentation describes the engine's capability to ingest and process external data such as RSS/Atom feeds, images, and favicons. While this constitutes an ingestion surface for indirect prompt injection, the skill itself is a static configuration guide and does not implement unsafe data handling logic.
[CREDENTIALS_UNSAFE]: The authentication section clearly documents a 'fail-closed' security model where the engine denies all access by default. It includes explicit warnings that the open_access setting should only be used in demo or non-production environments.

sm-configure