amq-cli

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.90). High risk: the SKILL instructs piping a raw GitHub-hosted install.sh from a personal repository to bash — a direct .sh download-and-execute from an unvetted user/repo is a common malware vector.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.70). SKILL.md instructs runtime use of amq env/amq who/amq drain-style commands that read queue/message contents (which can be authored by other agents/users) into the agent’s execution context, creating an indirect prompt-injection path via AMQ message bodies/frontmatter.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill's prerequisites explicitly run a remote install script via curl piped to bash (https://raw.githubusercontent.com/avivsinai/agent-message-queue/main/scripts/install.sh), which fetches and executes remote code as a required setup step.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.90). The skill explicitly advises running a remote installer via curl|bash and includes command-line flags (--dangerously-skip-permissions, --dangerously-bypass-approvals-and-sandbox) that bypass permissions/sandboxing, which push the agent toward altering system state and weakening security controls.