Security audit
Security checks for vulnerabilities and agentic risk
This looks like a legitimate business automation skill, but it gives broad command and scheduling guidance and includes under-controlled financial auto-approval examples.
Review this skill before installing if you might use it against real business systems. Keep it in design or dry-run mode until each workflow has explicit approval gates, audited thresholds, safe credential handling, rollback plans, and no silence-equals-approval behavior for payments, account changes, external messages, or production data updates.
The listed trigger keywords include generic terms such as "turn," "architect," "automation," "business," and "agent," which are common words that could appear in many unrelated user requests. The file does not provide narrowing constraints, explicit activation conditions, or negative examples to distinguish when this skill should activate versus when it should not.
The skill explicitly enables automation via exec, APIs, scripts, and cron-like workflows, but does not prominently warn users that it may perform system-impacting actions or external communications. In an agent environment with exec enabled, this can lead users to authorize or invoke automations that modify systems, schedule jobs, or transmit data without appreciating the operational risk.
The workflow pattern includes automatic approval of invoices within a threshold, which can authorize financial actions without human review if upstream matching or extraction is wrong. In a business automation skill, this is risky because invoice parsing errors, spoofed submissions, or weak PO matching could trigger unauthorized payments or fraud.
→ Extract data (vendor, amount, line items, due date)
→ Match to PO / budget category
→ Validate:
- Amount within approved range? → Auto-approve
- Over threshold? → Route to manager
- No matching PO? → Flag for review
→ Schedule payment based on terms
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
log "Fetching data from [source]..."
DATA=$(curl -s -H "Authorization: Bearer $API_TOKEN" \
"https://api.example.com/endpoint")
if [ -z "$DATA" ]; then
log "ERROR: No data returned"
The approval-gate example permits auto-approval on timeout for requests under a monetary threshold, which weakens the control by turning non-response into consent. This is dangerous because notification failures, missed messages, or deliberate timing abuse could cause unauthorized transactions or commitments to proceed without actual approval.
on_reject: "notify_requestor_with_reason" on_timeout: - Escalate to next level - Or: auto-approve if amount < $10000
### Graceful Degradation
The skill includes substantial Chinese-language content for core capability, scenarios, and trigger metadata, while the rest of the document is in English. There is no statement offering the user a language preference or clarifying that bilingual content is optional, which can create a language/locale policy issue.
No suspicious patterns detected.