Most static command findings are Markdown code fences, and the URL findings are documentation or in-cluster references, so those are false positives. The OpenTelemetry examples include all-interface listeners, anonymous Kiali access, and insecure TLS settings that require production hardening.
The Kiali example uses anonymous authentication and makes all namespaces accessible. Copying this into a shared cluster can expose service topology and traffic metadata without login controls.
The cited Kiali manifest explicitly sets anonymous auth and wildcard namespace access. The risk depends on deployment exposure, but the insecure default is clear.
The OpenTelemetry example sets tls.insecure to true for the Jaeger exporter. This disables transport verification in copied deployments unless limited to a trusted test network.
The Jaeger exporter configuration directly includes tls.insecure: true. The example may target a lab cluster, so confidence is high but not absolute.
Capability review items (4)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
The OpenTelemetry example binds a telemetry endpoint to 0.0.0.0, which listens on all interfaces if deployed. In production this can expose collectors unless network policy or authentication restricts access.
The OpenTelemetry example binds a telemetry endpoint to 0.0.0.0, which listens on all interfaces if deployed. In production this can expose collectors unless network policy or authentication restricts access.
The OpenTelemetry example binds a telemetry endpoint to 0.0.0.0, which listens on all interfaces if deployed. In production this can expose collectors unless network policy or authentication restricts access.
The OpenTelemetry example binds a telemetry endpoint to 0.0.0.0, which listens on all interfaces if deployed. In production this can expose collectors unless network policy or authentication restricts access.
Most static command findings are Markdown code fences, and the URL findings are documentation or in-cluster references, so those are false positives. The OpenTelemetry examples include all-interface listeners, anonymous Kiali access, and insecure TLS settings that require production hardening.
The Kiali example uses anonymous authentication and makes all namespaces accessible. Copying this into a shared cluster can expose service topology and traffic metadata without login controls.
The cited Kiali manifest explicitly sets anonymous auth and wildcard namespace access. The risk depends on deployment exposure, but the insecure default is clear.
The OpenTelemetry example sets tls.insecure to true for the Jaeger exporter. This disables transport verification in copied deployments unless limited to a trusted test network.
The Jaeger exporter configuration directly includes tls.insecure: true. The example may target a lab cluster, so confidence is high but not absolute.
Capability review items (4)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
The OpenTelemetry example binds a telemetry endpoint to 0.0.0.0, which listens on all interfaces if deployed. In production this can expose collectors unless network policy or authentication restricts access.
The OpenTelemetry example binds a telemetry endpoint to 0.0.0.0, which listens on all interfaces if deployed. In production this can expose collectors unless network policy or authentication restricts access.
The OpenTelemetry example binds a telemetry endpoint to 0.0.0.0, which listens on all interfaces if deployed. In production this can expose collectors unless network policy or authentication restricts access.
The OpenTelemetry example binds a telemetry endpoint to 0.0.0.0, which listens on all interfaces if deployed. In production this can expose collectors unless network policy or authentication restricts access.
The high static score is mostly false positives from Markdown code fences, PromQL windows, and documentation links. The skill is a Markdown guide with no executable package code, but several deployable examples use broad access or insecure defaults, so publication should include a security warning.
1
Files scanned
384
Lines analyzed
5
Review items
4
False positives ignored
Capability review items (3)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
The Kiali example uses anonymous authentication and access to all namespaces. If applied unchanged, it can expose mesh topology and telemetry beyond the intended audience.
The YAML directly shows anonymous auth and a wildcard namespace scope. The context is a template, so this is risky guidance rather than confirmed malicious behavior.
Open Telemetry Example Uses Broad Network Bindings
The collector example binds several receivers to 0.0.0.0 and disables TLS verification for the Jaeger exporter. This can expose telemetry endpoints in permissive networks.
The endpoint and TLS settings are explicit in the sample configuration. Risk depends on cluster network policy and whether users deploy the example unchanged.
The Linkerd example pipes generated manifests directly into kubectl apply. This is a common installation pattern, but users should review generated manifests before applying them.
The command clearly changes cluster state, but it uses the standard Linkerd CLI flow. I did not find evidence of hidden command execution.
Static false positives ignored (4)
These static matches were dismissed by semantic review or matched schema-only tokens, so they are shown for transparency but do not drive the quality score.
Markdown Backtick Detections Are Mostly False Positives
The static external command alerts mostly point to fenced Markdown examples, not executable Ruby or shell code. Only the Linkerd bash block contains user-run commands.
The flagged regions are Markdown code fences containing YAML, PromQL, JSON, and bash examples. There is no runnable skill script that automatically executes these snippets.
Weak Crypto Detections Are Query Syntax False Positives
The weak cryptography alerts match PromQL windows, 5xx patterns, quantiles, and Markdown text. I did not find MD5, SHA1, DES, or similar cryptographic use.
The flagged lines are observability metric expressions and descriptive text. No cryptographic API, hash algorithm selection, or key material handling appears in the file.
The high static score is mostly false positives from Markdown code fences, PromQL windows, and documentation links. The skill is a Markdown guide with no executable package code, but several deployable examples use broad access or insecure defaults, so publication should include a security warning.
1
Files scanned
384
Lines analyzed
5
Review items
4
False positives ignored
Capability review items (3)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
The Kiali example uses anonymous authentication and access to all namespaces. If applied unchanged, it can expose mesh topology and telemetry beyond the intended audience.
The YAML directly shows anonymous auth and a wildcard namespace scope. The context is a template, so this is risky guidance rather than confirmed malicious behavior.
Open Telemetry Example Uses Broad Network Bindings
The collector example binds several receivers to 0.0.0.0 and disables TLS verification for the Jaeger exporter. This can expose telemetry endpoints in permissive networks.
The endpoint and TLS settings are explicit in the sample configuration. Risk depends on cluster network policy and whether users deploy the example unchanged.
The Linkerd example pipes generated manifests directly into kubectl apply. This is a common installation pattern, but users should review generated manifests before applying them.
The command clearly changes cluster state, but it uses the standard Linkerd CLI flow. I did not find evidence of hidden command execution.
Static false positives ignored (4)
These static matches were dismissed by semantic review or matched schema-only tokens, so they are shown for transparency but do not drive the quality score.
Markdown Backtick Detections Are Mostly False Positives
The static external command alerts mostly point to fenced Markdown examples, not executable Ruby or shell code. Only the Linkerd bash block contains user-run commands.
The flagged regions are Markdown code fences containing YAML, PromQL, JSON, and bash examples. There is no runnable skill script that automatically executes these snippets.
Weak Crypto Detections Are Query Syntax False Positives
The weak cryptography alerts match PromQL windows, 5xx patterns, quantiles, and Markdown text. I did not find MD5, SHA1, DES, or similar cryptographic use.
The flagged lines are observability metric expressions and descriptive text. No cryptographic API, hash algorithm selection, or key material handling appears in the file.
Pure documentation skill containing YAML templates, PromQL queries, and CLI examples for service mesh observability. All static findings are false positives: the scanner misinterpreted PromQL metric names (containing 'md5', 'sha' substrings) as weak crypto, flagged documentation links as network IOCs, and misidentified YAML field names as path traversal. The content is static documentation that matches its stated purpose exactly.
Pure documentation skill containing YAML templates, PromQL queries, and CLI examples for service mesh observability. All static findings are false positives: the scanner misinterpreted PromQL metric names (containing 'md5', 'sha' substrings) as weak crypto, flagged documentation links as network IOCs, and misidentified YAML field names as path traversal. The content is static documentation that matches its stated purpose exactly.
Pure prompt-based skill containing static documentation and YAML/PromQL templates for service mesh observability. No code execution paths, no credential access, no file system or network operations. The content matches the stated purpose exactly.
Pure prompt-based skill containing static documentation and YAML/PromQL templates for service mesh observability. No code execution paths, no credential access, no file system or network operations. The content matches the stated purpose exactly.