The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill recommends a installation method that downloads a shell script from a remote URL and pipes it directly into the bash interpreter.
Evidence: curl -s https://raw.githubusercontent.com/koush/scrypted/main/install/docker/install-scrypted-docker-compose.sh | bash in SKILL.md.
[COMMAND_EXECUTION]: The skill requires and uses high-privilege system tools for installation, diagnostics, and monitoring.
Evidence: Commands such as npm install -g, docker logs, ffprobe, ping, and avahi-browse are referenced throughout the documentation for use with the Bash tool.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its handling of untrusted data from system logs and network streams.
Ingestion points: The skill reads external data via docker logs scrypted and network discovery tools like avahi-browse (SKILL.md).
Boundary markers: No delimiters or explicit instructions to ignore embedded commands are present in the provided instructions.
Capability inventory: The skill has access to Bash, Edit, Write, and WebFetch tools (YAML frontmatter).
Sanitization: There is no evidence of sanitization or validation for the content of logs or stream metadata before processing.

scrypted