Updated: 2026-03-25
security-testing
Basic Information
Full Skill Guide
When
- Sensitive data, auth, or transaction flows are introduced/changed.
- Compliance or internal policy requires security validation.
- Potential abuse vectors are identified in design review.
What
- Identify exploitable weaknesses in auth, input handling, and data protection.
- Assess security risk impact and remediation priority.
- Support secure-release decisions with evidence.
How
- Define threat scope and high-value assets.
- Build attack surface map across endpoints, roles, and trust boundaries.
- Execute checks for auth bypass, injection, data exposure, and misconfiguration.
- Validate logging, rate limits, and defensive controls.
- Reproduce vulnerabilities and estimate exploitability.
- Deliver risk-ranked findings with fix and revalidation guidance.
Reference
Positive Example (Input -> Output)
Input:
- Feature: admin export API with role-based access
Output:
- Found horizontal privilege escalation via parameter tampering
- Risk: high (sensitive data exposure)
- Fix: enforce server-side ownership check + add audit rule
Negative Example (Input -> Output)
Input:
- "Do a security scan"
Output (problem):
- Scanner results copied without exploit verification
- False positives mixed with real risks
Limits
- Do not treat scanner output as final verdict.
- Do not test production attack scenarios without authorization.
- Do not understate exploitability of data exposure issues.
- Do not omit reproduction details for confirmed findings.
- Do not sign off security status without remediation retest.
Usage Guide
- Install and enable
security-testingfirst (use the install commands in this page). - In your request, provide required context: scope, environment, timeline, and expected output format.
- Trigger with approved scope, for example: "Use
security-testingfor auth and sensitive-data APIs." - Ask for outputs: threat checklist, test plan, finding severity rules, and evidence format.
- After remediation, request focused revalidation steps for each high-risk finding.
Installation
Platform
AI Tool
Quick install (one line)
Generating command... Full script
Loading script...