- Home
- Skills
- Testing & Debugging
- Security Testing
Works with the AI Agents you already use
Security Testing
Prove a security control works instead of assuming it does
$19
Security Testing
Example session with this skill installed
Test the authorization control for the staging billing endpoint. A standard user must be able to read their own billing summary but must not access another tenant's billing data. Use the supplied test identities and return PASS, FAIL, or INCONCLUSIVE with evidence.
- Read your context and instructions
- Compiled the security testing
Security Control Test Record
CONTROL
A standard user may read their own billing summary but must not access another tenant's billing data.
TEST SCOPE
Synthetic staging billing endpoint with two controlled tenant identities.
EXPECTED ALLOWED BEHAVIOR
Tenant A user can access Tenant A billing summary.
EXPECTED DENIED BEHAVIOR
Tenant A user cannot access Tenant B billing summary.
Test ST-01
Objective
Verify the valid same tenant access path.
Synthetic observation
The worked example models the Tenant A identity receiving the expected Tenant A billing response.
Verdict
PASS IN SYNTHETIC EXAMPLE
The allowed path behaves as expected in the illustrative scenario.
Test ST-02
Objective
Verify the cross tenant negative case.
Synthetic observation
The worked example models the Tenant A identity attempting to request Tenant B billing data and being rejected by the authorization boundary.
Verdict
PASS IN SYNTHETIC EXAMPLE
The negative test demonstrates the expected denial behavior in the illustrative scenario.
Control Verdict
PASS IN SYNTHETIC EXAMPLE
Both the permitted path and the prohibited cross tenant path match the expected control behavior.
No real environment or production control is claimed as verified by this demo.
Evidence Boundary
Only the two modeled cases above are evaluated. No claim is made about other endpoints, roles, tenants, or authorization paths.
Next Step
For a real validation, execute the same positive and negative cases against the explicitly authorized environment and record the observed evidence. If required runtime evidence cannot be obtained, return INCONCLUSIVE instead of assuming PASS or FAIL.
Connects securely to your tools. The creator never sees your data.
About this skill
Plans and executes focused security tests against known controls and requirements, producing explicit PASS, FAIL, or INCONCLUSIVE evidence for each test objective.
What it does
- Converts security requirements and expected controls into focused test objectives.
- Runs static, dynamic, API, abuse case, fuzz, and configuration tests when applicable.
- Exercises negative cases to verify that prohibited behavior is actually rejected.
- Tests authorization, validation, boundary enforcement, and failure handling.
- Separates observed evidence from assumptions and unavailable runtime data.
- Returns an explicit verdict for every tested control.
- Records unresolved evidence gaps instead of converting them into false failures.
How it works
- Scope definition identifies the control, requirement, or expected security behavior to verify.
- Test design defines the expected allowed behavior and the corresponding negative case.
- Execution runs the smallest relevant test against the available code, configuration, interface, or authorized runtime.
- Evidence capture records the observable result needed to evaluate the control.
- Classification assigns PASS, FAIL, or INCONCLUSIVE based on that evidence.
- Handoff routes failed controls to remediation planning and unresolved exploitability questions to the appropriate specialist workflow.
Boundaries
This skill verifies known security controls and requirements.
Breadth focused discovery belongs to vulnerability-assessment. A single documented vulnerability that requires exploitability proof belongs to exploit-validation. Objective driven intrusion belongs to the appropriate penetration testing skill.
The skill does not turn every failed control into an exploitation campaign and does not treat missing evidence as proof that a control failed.
Why this beats prompting it yourself
Generic security testing prompts often mix discovery, exploitation, and verification into one workflow. They may also report a control as failed simply because the expected evidence was unavailable. This skill keeps the test objective explicit, exercises meaningful negative cases, records observable evidence, and distinguishes a real failure from an inconclusive result.
Validation
This skill is part of the JetFerrari Security Engineering stack, which passed pack level structural, routing, installation, discovery, and behavioral validation. No unsupported individual runtime qualification is claimed here.
Use cases
- Verify that an authorization control rejects forbidden access.
- Test a security requirement with explicit positive and negative cases.
- Validate input handling and security boundaries with focused fuzzing.
- Check whether configuration controls enforce the expected behavior.
- Produce PASS, FAIL, or INCONCLUSIVE evidence for release readiness.
Known limitations
Testing quality depends on the available implementation, configuration, runtime access, and observability. A control whose required evidence cannot be obtained remains INCONCLUSIVE. This skill verifies defined controls; it is not intended to discover every vulnerability or conduct an open ended penetration test.
Part of JetFerrari Security Engineering.
Security Testing verifies defined security requirements and controls through focused tests. It is not a breadth first vulnerability discovery workflow and does not perform chained penetration testing. Results depend on the available test surface, fixtures, identities, configuration, and runtime evidence, so insufficient execution context may produce INCONCLUSIVE rather than PASS or FAIL.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 24 days ago
Permissions
Access is limited to the security control and test scope supplied for the current task. The skill may inspect source files and configuration, run focused local tests, and interact with explicitly authorized application or API surfaces when required. It does not modify project files, harvest environment variables, expand into unrelated vulnerability discovery, or perform destructive testing.