ctf-sandbox-orchestrator

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly instructs recording requests/headers/cookies and "tokens, tickets" as evidence, which would require the LLM to handle or include secret values verbatim in outputs, creating an exfiltration risk.

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

• Third-party content exposure detected (low risk: 0.10). SKILL.md defines the “Competition Sandbox” as an orchestration layer that routes based on one minimal prompt-to-tool/observable runtime chain (and instructs treating “prompts, logs, HTML, JSON, comments, and docs” as untrusted artifacts), but the provided skill files do not describe any required runtime workflow that ingests outsider-authored free text from an external feed/queue/inbox/webpage without first selecting a specific target.