local-security

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt includes concrete examples of hardcoded API tokens (e.g., "ghp_abc123secrettoken") and shows embedding credentials directly into commands, which would require the model to output secret values verbatim and is therefore insecure.

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

• Attempt to modify system services in skill instructions detected (high risk: 0.80). The skill instructs the agent/user to modify local system state (generate SSH/GPG keys, change ~/.ssh/config, adjust file permissions, start/enable the ssh-agent service, and change credential helpers), including editing SSH config which is explicitly flagged as risky, so it pushes changes to the machine that could meaningfully alter security posture.