huawei-cloud-flexus-l-server-openclaw-deployment

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The prompt contains multiple non-interactive command examples and required parameters that embed AK/SK, security tokens, model API keys and channel secrets into CLI arguments or JSON (i.e., verbatim in generated commands/code), so an LLM would need to handle or output secret values directly despite recommending env vars.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The code contains high-risk remote-code-execution and credential-exposure patterns (curl|bash of remote scripts executed locally and via COC, creation+execution of arbitrary COC scripts on customer instances, and logging of Authorization/headers), which enable easy supply‑chain abuse, remote backdooring of target instances, and accidental or deliberate exfiltration of secrets.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 1.00). The skill creates and executes COC/local scripts that run remote shell installers via curl -sSL https://documentation-samples.obs.cn-north-4.myhuaweicloud.com/solution-as-code-publicbucket/solution-as-code-moudle/deploying-clawdbot-agents/userdata/multi_model.sh (and similar URLs multi_channel.sh, openclaw_gateway_manager.sh), which are fetched at runtime and piped to bash, so remote content directly executes code and is a required dependency for model/channel installation.

MEDIUM W021: Hidden or invisible Unicode characters detected (potential obfuscation or prompt injection).

• Hidden Unicode characters detected (1 type(s) found)