Security audit
Security checks for vulnerabilities and agentic risk
SkillGuard is a disclosed hosted security-audit skill that sends user-selected skill content to its API, with API-key setup and redaction guidance clearly documented.
Before installing, be aware that audits are performed by a hosted service: only submit skills and reference files you are allowed to share, redact secrets and proprietary material first, and use a least-privilege API key where possible.
The file’s user-facing natural-language instructions and descriptions are entirely in Chinese, including installation and operation guidance. This can violate a language/locale policy when the skill forces a specific language without user opt-in and does not state that the skill is intentionally limited to a Chinese-speaking or region-specific audience.
This markdown file is written entirely in Chinese and presents mandatory behavioral rules without any indication that the user may choose another language or that the restriction is required for a specific regional or compliance context. Under the policy, forcing a specific language without opt-in is a natural-language policy violation.
This documentation instructs users to transmit audit inputs to an external endpoint, including authorization credentials and the submitted skill content. External transmission is expected for a hosted audit API, but it still represents a real security/privacy risk if users are not clearly informed that sensitive local content will leave their environment.
API_ROOT="${AI_SKILLS_API_URL:-https://ai-skills.open-idea.net}/api/v1"
curl --fail-with-body "$API_ROOT/skillguard/audit" \
-H "Authorization: Bearer ${SKILLGUARD_API_KEY:-$AI_SKILLS_API_KEY}" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: audit-uuid-001" \
The example request sends full skill content and embedded file contents to a remote API, but the documentation does not prominently warn that proprietary code, prompts, secrets, or personal data may be transmitted off-host. In a security-audit context this is materially relevant because users may paste third-party skills for scanning and unintentionally disclose sensitive data to the service.
No suspicious patterns detected.