Security audit
Security checks for vulnerabilities and agentic risk
This is a disclosed paid translation skill that sends requested text to an external NEXUS API, with no evidence of hidden local access or persistence.
Install only if you trust NEXUS with the text you ask to translate and with the payment proof used for requests. Avoid sending secrets, regulated data, or proprietary text unless your organization allows that third-party processing, and be aware that non-sandbox use is priced per request.
Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.
Or manually copy the SKILL.md to your OpenClaw skills directory:
cp SKILL.md ~/.openclaw/skills/nexus-translate/SKILL.md
The README states the skill is 'automatically invoked' when a matching task is detected, but provides no concrete trigger boundaries, scoping rules, or user-consent requirements. In an agent ecosystem, ambiguous auto-invocation can cause sensitive user inputs to be routed to a paid external translation service unexpectedly, increasing the risk of unintended data disclosure and unauthorized spending.
The documented behavior sends user-provided input to an external host over the network for processing. For a translation skill this is functionally expected, but it still represents a real data-exfiltration risk if agents automatically forward sensitive prompts, credentials, proprietary text, or personal data to the third-party service without clear consent and policy controls.
# Step 1: Get the x402/MPP challenge
curl -X POST https://ai-service-hub-15.emergent.host/api/original-services/translate \
-H "Content-Type: application/json" \
-d '{"input": "your query here"}'
# Returns 402 + WWW-Authenticate: Payment header
The skill explicitly instructs agents to transmit user-provided text and a payment proof header to an external third-party service. Even though this is the stated purpose of a translation skill, it still creates a real data exfiltration/privacy boundary because potentially sensitive input is sent off-platform to a remote API and processed by upstream LLMs. The context makes this somewhat expected, but the combination of arbitrary input forwarding, external model processing, and payment-related headers means the transmission should be treated as a genuine security-relevant behavior rather than a false positive.
curl -X POST https://ai-service-hub-15.emergent.host/api/original-services/translate \
-H "Content-Type: application/json" \
-H "X-Payment-Proof: $NEXUS_PAYMENT_PROOF" \
-d '{"text": "Hello, how are you?", "target_language": "ja"}'
No suspicious patterns detected.