LVL 01
Easy
Reflected — Unfiltered
A search page echoes your query straight back into the HTML body. Get the page to run your script.
Not solved
Open →
Uncrypt // security playground
Learn web & AI security by breaking things — safely.
A hands-on lab environment for students and enthusiasts. Practice real attacks — XSS, SQL injection, CSRF, SSRF, open redirects, and modern AI/ML exploits — in isolated, intentionally-vulnerable sandboxes. No setup, no hints in the labs. Capture the flag, then read the walkthrough to understand exactly why it worked.
52
labs
7
attack tracks
0
solved by you
XSS
Cross-Site Scripting
Inject and execute script in another user's browser context.
0 / 10
✓ XSS Master earned
LVL 02
Easy
Reflected — Attribute Context
Your input lands inside an HTML attribute value. Break out and execute script.
Not solved
Open →
LVL 03
Medium
Reflected — JavaScript Context
Your input is placed inside an inline <script> string. Escape the string and run code.
Not solved
Open →
LVL 04
Medium
Stored — Guestbook
Comments are saved and shown to everyone who views the page. Plant a persistent payload.
Not solved
Open →
LVL 05
Medium
DOM — Hash Sink
Client-side JavaScript writes location data into the page. No server reflection involved.
Not solved
Open →
LVL 06
Medium
Filter — Tag Blacklist
The app strips the <script> tag. Achieve execution without it.
Not solved
Open →
LVL 07
Hard
Filter — Case & Encoding
A naive filter blocks some keywords. Slip past it.
Not solved
Open →
LVL 08
Medium
URL — javascript: Scheme
Your input becomes the target of a link. Turn the link into code execution.
Not solved
Open →
LVL 09
Medium
Event Handlers — img / svg
Angle brackets survive but the word "script" is blocked. Fire an event handler instead.
Not solved
Open →
LVL 10
Hard
Context — JSON / CSP Bypass
Your input is reflected into a JSON blob under a permissive policy. Achieve execution.
Not solved
Open →
SQLi
SQL Injection
Manipulate database queries through unsanitised input.
0 / 6
✓ SQLi Master earned
LVL 01
Easy
Authentication Bypass
A login form checks your credentials against the database. Log in as admin without the password.
Not solved
Open →
LVL 02
Medium
UNION-Based Extraction
A product lookup returns rows from the database. Extract data from another table.
Not solved
Open →
LVL 03
Medium
Error-Based
The application leaks database errors. Use them to pull data out.
Not solved
Open →
LVL 04
Hard
Blind — Boolean
The page only tells you "found" or "not found". Ask the database yes/no questions.
Not solved
Open →
LVL 05
Hard
Blind — Time-Based
No output, no errors, no difference — except how long the response takes.
Not solved
Open →
LVL 06
Hard
Filter Bypass — Keywords
Basic keyword filtering tries to block injection. Get around it.
Not solved
Open →
CSRF
Cross-Site Request Forgery
Force an authenticated user to perform unintended actions.
0 / 6
✓ CSRF Master earned
LVL 01
Easy
Change Email — No Token
A logged-in user can change their account email with no CSRF protection. Forge the request.
Not solved
Open →
LVL 02
Easy
State Change via GET
A sensitive action is performed over a simple GET request. Trigger it from off-site.
Not solved
Open →
LVL 03
Medium
Predictable Token
There is a CSRF token, but it is trivially guessable. Defeat it.
Not solved
Open →
LVL 04
Medium
Weak Referer Check
The server validates the Referer header — but poorly. Bypass the check.
Not solved
Open →
LVL 05
Medium
Login CSRF
Force the victim to log into an attacker-controlled account.
Not solved
Open →
LVL 06
Hard
Token Not Bound to Session
Tokens are validated but not tied to the user. Reuse one.
Not solved
Open →
SSRF
Server-Side Request Forgery
Make the server request internal resources on your behalf.
0 / 6
✓ SSRF Master earned
LVL 01
Easy
Basic — Internal Fetch
A URL-preview feature fetches any address you give it. Reach the internal admin panel.
Not solved
Open →
LVL 02
Medium
Cloud Metadata
The fetcher runs in the cloud. Retrieve the instance credentials.
Not solved
Open →
LVL 03
Hard
Blacklist Bypass
localhost and 127.0.0.1 are blocked. Reach the internal service anyway.
Not solved
Open →
LVL 04
Hard
Via Open Redirect
The fetcher refuses internal hosts, but follows redirects. Chain them.
Not solved
Open →
LVL 05
Hard
Blind SSRF
The response is hidden. Prove you reached the internal target.
Not solved
Open →
LVL 06
Medium
Internal Port Scan
Use the fetcher to discover what is listening on the internal host.
Not solved
Open →
Redirect
Open Redirect
Abuse redirect logic to send victims to attacker-controlled sites.
0 / 6
✓ Redirect Master earned
LVL 01
Easy
Basic — url Parameter
A "continue" link redirects wherever a parameter points. Send a victim off-site.
Not solved
Open →
LVL 02
Medium
Host-Check Bypass
The app checks that the target starts with the trusted host. Escape it.
Not solved
Open →
LVL 03
Medium
Meta / JS Redirect
The redirect happens client-side. Control the destination.
Not solved
Open →
LVL 04
Hard
Path Confusion
An allowlist compares against the wrong part of the URL. Exploit the parsing gap.
Not solved
Open →
LVL 05
Hard
Fragment Token Theft
A redirect carries a token in the URL fragment. Steal it via the redirect.
Not solved
Open →
LVL 06
Medium
Allowlist Suffix Trick
Only *.uncrypt.io is allowed. Register a domain that slips through.
Not solved
Open →
Modern
Modern Web Attacks
IDOR, injection, traversal, SSTI, JWT, uploads and logic flaws.
0 / 8
✓ Web Exploitation Master earned
LVL 01
Easy
IDOR — Broken Access
Invoices are addressed by a numeric id. Read one that is not yours.
Not solved
Open →
LVL 02
Medium
Command Injection
A network tool pings a host you supply. Run your own command.
Not solved
Open →
LVL 03
Medium
Path Traversal / LFI
A file viewer serves pages by name. Escape the directory.
Not solved
Open →
LVL 04
Hard
Server-Side Template Injection
A greeting template renders your name. Inject template syntax.
Not solved
Open →
LVL 05
Hard
JWT — alg Confusion
The session is a JWT. Forge an admin token.
Not solved
Open →
LVL 06
Hard
Insecure File Upload
An avatar uploader accepts your file. Get it treated as code.
Not solved
Open →
LVL 07
Hard
XXE — XML Parsing
A parser reads XML you submit. Make it read a local file.
Not solved
Open →
LVL 08
Medium
Mass Assignment / Privesc
A profile-update endpoint trusts your fields. Make yourself an admin.
Not solved
Open →
AI Sec
AI / ML Security
Prompt injection, jailbreaks, data leakage and model evasion.
0 / 10
✓ AI Security Master earned
LVL 01
Easy
Direct Prompt Injection
A support chatbot guards a secret. Talk it into revealing it.
Not solved
Open →
LVL 02
Medium
Indirect Prompt Injection
The assistant summarises a document you provide. Hide instructions inside it.
Not solved
Open →
LVL 03
Medium
System Prompt Leak
Recover the hidden system prompt the assistant was configured with.
Not solved
Open →
LVL 04
Medium
Jailbreak — Guardrail Bypass
The model refuses restricted requests. Get it to comply anyway.
Not solved
Open →
LVL 05
Hard
Insecure Output Handling
The app renders the model's answer as HTML. Turn a reply into XSS.
Not solved
Open →
LVL 06
Hard
Prompt Injection → Tool Abuse
The agent can call tools. Trick it into calling one it should not.
Not solved
Open →
LVL 07
Medium
Training-Data / PII Extraction
The model has memorised sensitive records. Extract one.
Not solved
Open →
LVL 08
Hard
Excessive Agency — SSRF via AI
The agent can fetch URLs. Point it at the internal metadata service.
Not solved
Open →
LVL 09
Medium
Adversarial Evasion
A toxicity classifier blocks certain messages. Get a blocked message through.
Not solved
Open →
LVL 10
Hard
Data Poisoning
A spam filter learns from user feedback. Poison it into misclassifying.
Not solved
Open →
About the playground
Built to be attacked — in the right places only.
Uncrypt Playground is a self-contained training ground. Each lab is an isolated, deliberately vulnerable mini-app. The platform around it — this site, your progress, the walkthroughs — is hardened and holds no real accounts or data, so there is nothing here worth stealing. Attack the labs; the playground itself stays out of scope.
01 — Capture flags
Exploit, don't guess
Every lab hides a flag like UNCRYPT{…} that only appears once you actually pull off the
attack. Submit it on the lab page to mark it solved.
02 — Earn badges
Clear a whole track
Finish every level in a category and you'll be prompted for a name and a downloadable badge — XSS Master, AI Security Master, and more. Share it and tag #UncryptPlayground.
03 — Understand it
No hints in the labs
Labs never spoil the solution. When you're ready, the Walkthroughs tab explains each bug end-to-end, with the payload and how to fix it.
04 — Stay safe
Sandboxed by design
Dangerous primitives (SSRF, command execution, file reads) run against bundled mocks — never the real host or network — so nothing you do here can escape the sandbox.
Progress is stored only in this browser. Reset my progress.