Security audit
Security checks for vulnerabilities and agentic risk
This skill does what it claims by sending text to external AI services for rewriting, so users should avoid submitting confidential material.
Install this only if you are comfortable sending submitted text to Claw0x and, on the LLM path, Gemini. Do not use it for secrets, regulated data, confidential drafts, or private customer information unless that external processing is approved.
The skill declares use of environment secrets and external network access but does not define explicit tool scope or allowed-tools constraints. That increases the chance an agent platform will invoke it with broader-than-necessary capabilities or without clear review of its data egress behavior.
The trigger phrases are broad and include common requests like 'make this sound more natural' or 'clean up robotic-sounding content,' which can cause unintended invocation. In context, unintended invocation matters because the skill sends user text to an external service and is explicitly positioned to help evade AI detection, so users may disclose sensitive content without realizing it leaves the platform.
The description and usage guidance do not clearly warn at the point of use that the submitted text is transmitted to a third-party API for processing. This creates a privacy and confidentiality risk because users may provide proprietary, personal, or regulated content under the assumption that rewriting happens locally.
This skill performs an external POST request containing arbitrary user-provided text to claw0x.com, creating direct data exfiltration from the agent environment to a third party. The risk is amplified by the skill's purpose: it encourages submission of drafts, emails, documentation, and other potentially sensitive text, and also notes server-side use of another LLM provider, expanding the processing chain.
curl -s -X POST https://claw0x.com/v1/call \
-H "Authorization: Bearer $CLAW0X_API_KEY" \
-H "Content-Type: application/json" \
-d '{
The handler sends user-supplied text to Google's Gemini API in humanizeWithLLM, but this file contains no explicit disclosure, consent flow, or control to prevent unexpected third-party data sharing. That is a real privacy/security issue because users may submit sensitive or proprietary text assuming local processing, and the skill’s purpose—rewriting text to avoid AI-detection—can encourage submission of confidential drafts, academic work, or workplace materials.
Detected: suspicious.env_credential_access