Most static findings are false positives from Markdown code fences, reserved example URLs, defensive samples, and standard local report processing. The CI template executes an unverified remote installer and uses mutable third-party actions in security-sensitive workflows. The skill also provides actionable IDS and WAF evasion guidance despite repeated authorization warnings.
The workflow downloads a script from a mutable branch and pipes it directly to Bash. A compromised upstream response would execute arbitrary code on the CI runner.
The skill teaches random user agents, URI mutations, IDS evasion, and WAF bypass techniques that can conceal unauthorized vulnerability scanning.
The headings and commands explicitly describe evading detection and continuing when a WAF blocks scans. Authorization warnings reduce misuse risk but do not remove the capability.
The CI template uses third-party actions through mutable tags, including a master branch, allowing upstream changes to execute in security-sensitive workflows.
The template visibly selects gitleaks-action@v2 and trivy-action@master instead of immutable commit SHAs. These actions run on CI runners with repository context.
Capability review items (1)
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 URL references an installer on the mutable master branch and its response is executed without checksum or signature verification. This creates a concrete supply-chain dependency.
Most static findings are documentation examples, Markdown formatting, or security education snippets rather than executable threats. I confirmed a critical CI pipe-to-shell installer, a related unpinned remote download, and risky GITHUB_TOKEN exposure. I also found semantic risk in the main skill guidance for stealth and IDS evasion during web scanning.
The CI template pipes an unpinned remote GitHub installer directly into bash. This creates remote code execution risk if the download path or upstream script is compromised.
The skill includes an Evasion and Stealth section with random user agents, random request injection, IDS evasion options, cookies, and WAF blocking workarounds. This can help users avoid detection during web scanning, even though the text says authorized testing only.
The section headings and commands directly describe evasion and WAF blocking workarounds in the main skill instructions. Authorization warnings reduce malicious intent evidence but do not remove the abuse-enablement risk.
Capability review items (2)
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 workflow passes GITHUB_TOKEN to a third-party action. This is a common CI pattern, but it exposes a powerful repository token to external action code and should be tightly pinned and scoped.
The hardcoded URL is not just a reference; it downloads an installer that is immediately executed in CI. The source is remote and unpinned, so network retrieval affects build integrity.
The audit confirms the CI template remote installer piped to bash as a critical supply-chain pattern and flags stealth guidance as a high-risk semantic issue. Most other static detections are false positives from Markdown examples, reference links, dummy credentials, and secure coding examples. No prompt injection attempt was found in the reviewed files.
Line 240 pipes a remote installer from raw.githubusercontent.com directly into bash. That allows upstream or transit compromise to execute code in CI without pinning or checksum verification.
The skill includes explicit evasion and stealth commands for Nikto, including IDS evasion options and WAF-blocking workarounds. This can enable unauthorized scanning to avoid detection despite the authorization warning.
The cited sections use clear evasion language and concrete command options. The authorization caveat reduces certainty about malicious intent but not misuse potential.
Capability review items (1)
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 URL is used to fetch an installer during CI as part of a curl-to-bash sequence. Runtime network dependency on unverified remote code is a supply-chain risk.
Static false positives ignored (2)
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.
Static analysis found many command, network, script, environment, and filesystem patterns. Most are false positives from documentation templates or expected Nikto usage, but the skill contains confirmed dual-use scanning guidance, IDS/WAF evasion examples, and a CI template that pipes a remote installer into a shell. No evidence found of prompt injection, credential exfiltration, or malicious payloads, so this is not blocked as malware.
Authorized Web Vulnerability Scanning Can Affect Targets
The skill is built around running Nikto against web servers. It warns users to get written permission and notes traffic, alerting, log volume, and possible production impact.
The intended functionality is active web server scanning, which inherently sends network traffic to targets. The documented authorization checks show legitimate intent, so this is elevated dual-use risk rather than confirmed abuse.
Capability review items (7)
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 skill gives explicit Nikto evasion and WAF-blocking workarounds. Even with authorization warnings, this materially increases dual-use risk because it helps bypass detection during web scanning.
The file explicitly labels an evasion section and shows commands for IDS evasion and WAF blocking scenarios. The authorization framing lowers malicious-intent confidence but not the dual-use risk.
The CI template demonstrates downloading a remote shell installer and piping it directly to bash. This is a supply-chain risk if copied because remote content executes inside CI without pinning or verification.
The pattern is direct remote code execution in a CI job. It appears instructional rather than malicious, but it remains an unsafe pattern for marketplace users.
Authenticated scanning examples place usernames, passwords, and cookies directly in command lines. These are placeholders, but copied patterns can leak credentials through shell history, logs, or process listings.
The examples are clearly placeholders for Nikto usage, not credential theft. The risk is still credible because users may copy the command structure with real secrets.
The skill shows multi-host and bulk scan workflows. This is useful for authorized assessment, but it can create high traffic or scan unintended systems if scope controls are weak.
The commands directly support scanning multiple targets. Authorization language exists elsewhere, so the concern is operational risk and misuse potential rather than malicious intent.
Environment Token Access Is Limited to Security Tool Configuration
The GitHub token and API key references are used in CI or secure-code examples. No evidence found that the skill exfiltrates secrets or instructs users to reveal tokens.
The static patterns are real environment access references, but semantic context shows secrets scanning and secure API examples. There is no network exfiltration path tied to these values.
Static hits for XSS, weak cryptography, hardcoded secrets, and injection occur inside rule templates or reference examples that show vulnerable and fixed patterns for education.
The files label these blocks as examples and include remediation or fixed versions. The content can still be copied incorrectly, but it is not executable skill behavior.
Local File Reads Are for Generated Report Comments
The Node.js filesystem access in the CI template reads a locally generated markdown report before creating a pull request comment. No evidence found of arbitrary file reads.
The filename is hardcoded to the report generated earlier in the same job. The pattern is legitimate CI automation, although it remains a filesystem access risk factor.
Documentation-only skill providing guidance for the Nikto open-source web server scanner. All code examples are shell commands for the external Nikto tool which users install and run separately. Contains no executable scripts. Includes explicit authorization requirements and ethical usage guidelines. The static findings (external commands, network URLs, security vocabulary) are expected documentation content for a vulnerability scanning skill, not malicious patterns.
6
Files scanned
2,216
Lines analyzed
4
Review items
0
False positives ignored
Capability review items (1)
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 skill documentation includes examples of Nikto scanning commands that interact with external web servers. These are documented usage instructions for a legitimate open-source security tool. Includes explicit authorization requirements and ethical usage guidelines.
Documentation-only skill providing guidance for the Nikto open-source web server scanner. All code examples are shell commands for the external Nikto tool which users install and run separately. Contains no executable scripts. Includes explicit authorization requirements and ethical usage guidelines. The static findings (external commands, network URLs, security vocabulary) are expected documentation content for a vulnerability scanning skill, not malicious patterns.
6
Files scanned
2,216
Lines analyzed
4
Review items
0
False positives ignored
Capability review items (1)
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 skill documentation includes examples of Nikto scanning commands that interact with external web servers. These are documented usage instructions for a legitimate open-source security tool. Includes explicit authorization requirements and ethical usage guidelines.
Documentation-only skill providing guidance for the Nikto web server scanner. Contains no executable scripts. All code examples are for the external Nikto tool which users install separately. The skill includes explicit authorization requirements and ethical usage guidelines.
5
Files scanned
1,957
Lines analyzed
3
Review items
0
False positives ignored
Capability review items (1)
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 SKILL.md documentation includes extensive examples of Nikto scanning commands that interact with external web servers. These are documented usage instructions for the Nikto tool, not executable code. An attacker could theoretically use this documentation to learn scanning techniques, but the documentation explicitly requires authorization before use.
Documentation-only skill providing guidance for the Nikto web server scanner. Contains no executable scripts. All code examples are for the external Nikto tool which users install separately. The skill includes explicit authorization requirements and ethical usage guidelines.
5
Files scanned
1,957
Lines analyzed
3
Review items
0
False positives ignored
Capability review items (1)
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 SKILL.md documentation includes extensive examples of Nikto scanning commands that interact with external web servers. These are documented usage instructions for the Nikto tool, not executable code. An attacker could theoretically use this documentation to learn scanning techniques, but the documentation explicitly requires authorization before use.
Documentation-only skill providing guidance for the Nikto web server scanner. Contains no executable scripts. All code examples are for the external Nikto tool which users install separately. The skill includes explicit authorization requirements and ethical usage guidelines.
5
Files scanned
1,957
Lines analyzed
3
Review items
0
False positives ignored
Capability review items (1)
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 SKILL.md documentation includes extensive examples of Nikto scanning commands that interact with external web servers. These are documented usage instructions for the Nikto tool, not executable code. An attacker could theoretically use this documentation to learn scanning techniques, but the documentation explicitly requires authorization before use.