Security audit
Security checks for vulnerabilities and agentic risk
The skill is mostly a coherent Huawei Cloud observability troubleshooting guide, but it includes cloud resource creation steps inside the diagnostic flow without clear approval or cost warnings.
Install only if you are comfortable giving the agent Huawei Cloud observability and IAM read access, and do not let it run CreateBusiness, CreatePromInstance, or other subscription/resource-creation steps unless an authorized user explicitly approves the change and cost impact. Replace cn-north-4 with the correct region for your environment.
The skill is presented as a diagnosis workflow, but it includes write operations that create APM and AOM resources during troubleshooting. That breaks the expected read-only safety boundary for a diagnostic skill and can cause unintended infrastructure changes, billing impact, and configuration drift if an agent or user follows it automatically.
The documented diagnosis steps instruct users to create cloud resources without prominently warning that these actions modify the environment and may incur charges. In an agent skill context, omission of side-effect and cost warnings increases the chance of unsafe or unauthorized changes being made under the guise of troubleshooting.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
4. **agent-ops query side** (API only — not in KooCLI):
```bash
curl -X POST ${AGENT_OPS_ENDPOINT}/v1/ops/observation/traces \
-H "Authorization: Bearer ${IAM_TOKEN}" \
-H "Content-Type: application/json" \
-d '{"start_time": 1720000000000, "end_time": 1720600000000}'
The manifest describes a skill focused on diagnosing missing metrics, logs, traces, sessions, subscription metadata, data pipelines, query parameters, and authentication in the observability stack. Including generic IAM policy listing and authorization schema discovery adds a broader identity-enumeration capability that is not clearly justified by the stated diagnosis purpose, especially in a static test template file.
The document requires verification of timestamp formatting and region consistency, which may imply fixed locale or regional constraints in diagnostic handling. However, the file does not clearly state whether these constraints are user-selectable or justified as part of a region-specific workflow.
The report template fixes the Region field to cn-north-4, which is a locale/region constraint expressed in natural language content. Because the template does not say this is optional, user-selected, or limited to a region-specific workflow, it can steer users toward a specific locale without opt-in.
The command examples all include --cli-region=cn-north-4, which imposes a specific locale/region in the template content. The file does not offer an alternative region selection or explain that the commands are valid only for a region-specific environment, so this is a natural-language locale policy concern.
The example command fixes --cli-region=cn-north-4, and the same pattern appears elsewhere in the file. For a general IAM policy reference, this imposes a locale/region-specific default without documenting that the skill is intended only for that region or giving users an opt-in choice.
The command example mandates --cli-region=cn-north-4 with no note that the file is region-scoped. This can violate locale/region policy expectations when the skill appears to be generic but only demonstrates one fixed region.
This example again uses --cli-region=cn-north-4 as a fixed region while the surrounding documentation does not explain a region limitation. A generic reference should either justify the regional constraint or allow user choice.
The system policy verification command fixes the CLI region to cn-north-4 without any stated justification or opt-in. This is a natural-language/configuration policy issue because the file otherwise reads like a general IAM reference.
Multiple command templates force --cli-region=cn-north-4, which is a locale-specific constraint embedded throughout the file. Because the file does not indicate that this region is optional, user-selectable, or justified by a region-specific purpose, it risks violating language/locale policy expectations.
No suspicious patterns detected.