The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill performs piped remote code execution by downloading a shell script from https://strix.ai/install and piping it directly to bash. This is implemented in scripts/install.sh and described as the primary setup method in SKILL.md, SKILL.toon, and references/commands.md.
[EXTERNAL_DOWNLOADS]: The skill downloads binaries and scripts from strix.ai and pulls container images from the GitHub Container Registry (ghcr.io/usestrix/strix-sandbox), a well-known service.
[COMMAND_EXECUTION]: The skill relies on the Bash tool to execute operational scripts including scripts/install.sh, scripts/run-scan.sh, and scripts/ci-scan.sh, which in turn manage the strix CLI and docker environment.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface through its processing of untrusted targets and instruction files. 1. Ingestion points: Untrusted data enters via the --target, --instruction, and --instruction-file parameters in scripts/run-scan.sh and scripts/ci-scan.sh. 2. Boundary markers: The skill does not implement delimiters or protective instructions to isolate potentially malicious embedded instructions in target codebases. 3. Capability inventory: The skill can execute shell commands, perform network operations, and access local files through the strix CLI. 4. Sanitization: No input validation or sanitization is applied to arguments before they are passed to the shell for execution.

strix