Most static hits are false positives from Markdown syntax, reference links, and labeled security examples. The CI template has a critical supply-chain risk because it pipes an unpinned remote script into Bash. The skill also provides high-impact credential fuzzing and reconnaissance workflows that require explicit authorization.
The CI template pipes an unpinned remote installer directly into Bash without checksum or signature verification. Upstream compromise would produce arbitrary runner execution.
The recommended reconnaissance workflow includes nmap port scanning. It is dual-use active network discovery that requires explicit target authorization.
Credential Fuzzing and Account Enumeration Workflow
The skill instructs users to identify valid usernames, test password lists, and combine credential lists. This can enable account compromise outside an authorized scope.
The workflow explicitly describes username discovery, password fuzzing, and clusterbomb credential combinations against login endpoints.
-mr "Invalid password|Incorrect password" # Match responses indicating valid user
The response matcher is expressly used to identify valid usernames from authentication differences. That creates a real account-enumeration capability.
The prose explicitly advises avoiding detection while fuzzing. Rate limiting reduces load, but the stated evasion goal can also support stealthy unauthorized scanning.
Discovery and Retrieval of Exposed Sensitive Files
The workflow searches for backup and sensitive files, then advises downloading discovered content. This can expose credentials, personal data, or proprietary information.
The cited workflow explicitly enumerates backup extensions and sensitive paths, then directs the user to download and analyze discovered files.
-fs 0 # Filter by response size to identify valid vhosts
The ffuf command performs virtual-host enumeration by comparing responses. This is legitimate authorized DAST functionality but remains an active reconnaissance capability.
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 points to a mutable branch script whose response is executed immediately. A compromised upstream branch or connection can deliver arbitrary CI code.
Most static alerts are Markdown examples, placeholder URLs, or reference text, but the review confirms a critical CI supply-chain issue and high-risk dual-use DAST guidance. The skill is useful for authorized application security testing, yet it can enable reconnaissance, credential enumeration, and sensitive file discovery if used outside scope.
-mr "Invalid password|Incorrect password" # Match responses indicating valid user
The line matches login responses that indicate valid users during username fuzzing. That can enable account enumeration if used outside authorized testing.
The CI template uses third-party actions by mutable version tags and includes a remote installer path. This can expose users to supply-chain changes when the template is copied into a workflow.
The file shows gitleaks and trivy actions referenced by tag or branch, plus a remote installer. These are clear CI supply-chain risks when reused without pinning or integrity checks.
Credential Fuzzing Against Authentication Endpoints
The skill describes fuzzing usernames and passwords against login endpoints, including matching responses that indicate valid users. This is dual-use and can enable account enumeration or brute-force activity outside an approved test.
The authentication workflow explicitly covers username discovery, password fuzzing, and combined username-password fuzzing. The behavior is legitimate for authorized DAST but high risk outside scope.
Sensitive Backup and Data Exposure Discovery Workflow
The skill guides users to find backup files, sensitive files, and exposed data, then download and analyze discovered files. This can lead to unauthorized data access when used outside a permitted scope.
The workflow directly targets backup and sensitive files and instructs users to analyze discovered files. The surrounding guidance includes authorization warnings, so the risk is dual-use rather than clearly malicious.
The section says to rate-limit fuzzing to avoid detection while tuning delays and concurrency. That wording can support evasive use despite also reducing traffic impact.
The rate-limited fuzzing section frames delays and concurrency limits as a way to avoid detection. This wording can encourage evasive scanning behavior instead of only safety-focused traffic reduction.
The phrase appears in a section about tuning ffuf request pacing. It has a benign performance interpretation, but the detection-avoidance wording is materially risky.
-fs 0 # Filter by response size to identify valid vhosts
The line supports virtual-host discovery by filtering response sizes to identify valid hosts. This is legitimate in scope, but it is still reconnaissance against web targets.
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 CI template passes GITHUB_TOKEN into a third-party action. This is common for GitHub workflows, but it creates token exposure risk if the action or ref is compromised.
The URL is part of a remote installer fetched by curl and piped to bash. A hardcoded remote script source creates supply-chain exposure if the template is used.
I confirmed the remote pipe-to-shell CI installer and the intended active web-fuzzing workflows as security-relevant risks. Most static hits are false positives from Markdown code fences, placeholder URLs, security examples, and generic templates. No prompt injection attempt was found in the reviewed files.
The template uses a curl-to-bash installer pattern for tfsec. Executing network content directly in CI can run attacker-controlled code if the upstream script or transport path is compromised.
-mr "Invalid password|Incorrect password" # Match responses indicating valid user
The authentication workflow matches responses that reveal valid usernames. This can enable credential enumeration if run outside an approved assessment.
The integration workflow explicitly includes nmap port scanning before ffuf enumeration. This is an active reconnaissance capability that requires strict authorization.
The skill guides directory, virtual host, parameter, and sensitive file discovery against target web applications. This is legitimate DAST only when the target is explicitly authorized.
The workflows explicitly describe content discovery, vhost identification, and sensitive file discovery. The file also includes authorization guidance, so the finding is about misuse risk rather than hidden malicious intent.
Credential Enumeration and Password Fuzzing Workflow
The authentication workflow directs username discovery and password fuzzing against login endpoints. This can enable credential enumeration or account lockouts outside a controlled assessment.
The workflow explicitly fuzzes usernames and passwords and matches responses indicating valid users. That is a clear authentication abuse risk if authorization and safeguards are absent.
-fs 0 # Filter by response size to identify valid vhosts
The workflow identifies valid virtual hosts by comparing responses. This is legitimate DAST when authorized, but it is still reconnaissance against web infrastructure.
The wording mentions avoiding detection while tuning scan rate limits. The context also discusses reducing load, so confidence is moderate, but the phrase can support stealthy reconnaissance.
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 CI template downloads an installer from GitHub and pipes it directly to bash. This is a real supply-chain risk because remote script content is executed without pinning or checksum verification.
The skill includes a CI shell script that runs ffuf and parses scan output. This is expected for the tool, but it is real command execution that should remain scoped and user-approved.
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 reported many command, network, environment, script, and filesystem patterns, but review shows most are documentation examples or security-rule templates. The skill is still medium risk because it provides real ffuf network fuzzing and credential-fuzzing workflows that require authorization and rate limits. No prompt injection or confirmed malicious intent was found.
The skill documents real ffuf workflows for directory discovery, parameter fuzzing, virtual host discovery, and CI testing. This is appropriate for authorized DAST work, but the same commands can be misused for reconnaissance against unauthorized targets.
The command examples and the safety section confirm legitimate security-testing intent. Risk remains elevated because the network fuzzing behavior is directly usable outside authorized scope.
The skill includes examples for fuzzing usernames and passwords against login endpoints. This is legitimate in a scoped test, but it is sensitive because it can resemble credential stuffing or brute-force activity if used without permission.
The examples explicitly target login requests and password wordlists. The surrounding DAST context reduces malicious-confidence, but the operational risk is clear.
The CI template installs tfsec by piping a remote script directly into bash. This is a real supply-chain risk if copied as-is because remote script content is executed without pinning or verification.
The line contains a direct remote download piped to bash. It appears to be a template example, not malicious code, but the execution pattern is dangerous.
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.
Many external command and URL findings are expected markdown examples for installing ffuf and running authorized web fuzzing. They are not hidden automation and do not execute when the skill is loaded.
The patterns are in fenced documentation blocks and match the stated purpose of an ffuf DAST skill. No evidence found that the skill silently executes them.
The rule template includes sample vulnerable code for hardcoded secrets, XSS, and weak cryptography so security rules can detect those patterns. These examples are educational content rather than active malicious behavior.
The surrounding metadata labels the patterns as detection rules and remediation examples. No evidence found that these examples are executed by the skill.
Environment Token Use Is GitHub Action Configuration
The GitHub token reference is used by the Gitleaks action in a CI template. This is normal GitHub Actions configuration, not evidence of credential exfiltration.
The token is referenced through GitHub Actions secrets and passed to a named secrets-scanning action. No external exfiltration endpoint is shown.
XSS and API Key Examples Are Remediation Documentation
The example document shows unsafe DOM operations and environment-based API key handling as part of security education. These snippets are not loaded by an application runtime in this skill.
The sections are labeled as vulnerability examples or secure API authentication examples. No evidence found of active script execution or secret theft.
Documentation-only skill providing guidance for ffuf, a legitimate open-source DAST tool. All static findings are in markdown files and YAML templates showing example commands. No executable code, network operations, or credential access present in the skill itself. Heuristic alerts trigger on expected DAST tool patterns (command execution, network requests, credential handling) which are legitimate functionality for web fuzzing.
Documentation-only skill providing guidance for ffuf, a legitimate open-source DAST tool. All static findings are in markdown files and YAML templates showing example commands. No executable code, network operations, or credential access present in the skill itself. Heuristic alerts trigger on expected DAST tool patterns (command execution, network requests, credential handling) which are legitimate functionality for web fuzzing.
Documentation-only skill containing no executable code. Provides guidance for using the legitimate open-source ffuf security tool. All files are template documentation with defensive security patterns. No network operations, file system access, or code execution paths present.
5
Files scanned
1,896
Lines analyzed
0
Review items
0
False positives ignored
No confirmed security findings were recorded for this completed audit.
Documentation-only skill containing no executable code. Provides guidance for using the legitimate open-source ffuf security tool. All files are template documentation with defensive security patterns. No network operations, file system access, or code execution paths present.
5
Files scanned
1,896
Lines analyzed
0
Review items
0
False positives ignored
No confirmed security findings were recorded for this completed audit.
Documentation-only skill containing no executable code. Provides guidance for using the legitimate open-source ffuf security tool. All files are template documentation with defensive security patterns. No network operations, file system access, or code execution paths present.
5
Files scanned
1,896
Lines analyzed
0
Review items
0
False positives ignored
No confirmed security findings were recorded for this completed audit.