ClawAudit verdict
trtc-ai-customer-service
Tencent TRTC-based AI customer service app builder using Flask and WebRTC; all API calls go to documented Tencent Cloud endpoints, credentials are user-configured TRTC keys used for stated purpose.
⚠ Flagged for review — coarse, uncorroborated signal, not a confirmed exploit. Review the config yourself before installing.
Automated static analysis — not a human review. ClawAudit flags capabilities, not confirmed intent, and can produce false positives. Disagree with this verdict? Use Dispute below.
Findings (9)
<script> tag in markdown — potential code injection
references/frontend-guide.md · code · <script
Pipe to python — executes piped content as Python code
SKILL.md · prose · downgraded · | Python
Possible hardcoded credential
assets/start.sh · prose · downgraded · SECRET : " ; read -r INPUT_TSECRET </dev/tty 2>/dev/null || INPUT_TSECRET=
Attempts to modify skills or system prompts
assets/start.sh · prose · downgraded · edit SystemPrompt
References sudo — requests elevated privileges
SKILL.md · prose · downgraded · sudo
Instructs covert action — may act without user awareness
assets/start.sh · prose · downgraded · silently
References agent configuration files
references/architecture.md · code · AgentConfig
HTTP request to bare IP address — common in malicious payloads
scripts/scaffold.py · prose · downgraded · https://0.0.0.0
Base64 encoding/decoding
assets/TLSSigAPIv2.py · prose · downgraded · base64_encode
Permissions & capabilities
Requires 1 system binary.
Is this flag fair?
Thanks — recorded.