| name | secureReview |
| description | Use when: reviewing code for OWASP Top 10:2025 security findings with severity classification and a structured output format. |
| type | reference |
| version | 1.0 |
| license | MIT |
Secure Review
Skill metadata: version "1.0"; tags [security, owasp, review]; recommended tools [].
Use this skill in workspaces with the secure pack selected.
Perform security-focused code review against the OWASP Top 10:2025. For each finding, map it to a specific OWASP category.
When to use
- Performing security-focused code review in workspaces with the secure pack selected
When NOT to use
- When reviewing infrastructure or deployment code — prefer
devopsReview
- When dependency scanning is the primary goal — prefer
dependencyAudit
OWASP Top 10:2025 categories
| ID | Category | Common patterns to look for |
|---|
| A01 | Broken Access Control | Missing authorization checks, IDOR, path traversal, privilege escalation |
| A02 | Security Misconfiguration | Default credentials, overly permissive CORS, debug endpoints in production, unnecessary features enabled |
| A03 | Software Supply Chain Failures | Unpinned dependencies, unsigned artifacts, build system trust issues |
| A04 | Cryptographic Failures | Plaintext sensitive data, weak algorithms (MD5/SHA1 for secrets), hardcoded keys, insufficient key length |
| A05 | Injection | SQL injection, command injection, LDAP injection, template injection, XSS |
| A06 | Insecure Design | Missing rate limiting, no fraud controls, unsafe business logic, design decisions that assume trust |
| A07 | Authentication Failures | Weak password policy, missing MFA, session fixation, predictable session tokens |
| A08 | Software or Data Integrity Failures | Deserialization of untrusted data, unsigned updates, missing integrity checks |
| A09 | Security Logging and Alerting Failures | No audit trail, logging sensitive data, no alerting on critical events |
| A10 | Mishandling of Exceptional Conditions | Stack traces exposed to users, unsafe exception paths, resource leaks on error |
Review output format
Produce a table for every code review:
| Severity | OWASP | File:Line | Finding | Recommended fix |
|---|
- Severity: Critical / High / Medium / Low
- OWASP: category ID (A01–A10) or "N/A" if not an OWASP category
- If zero findings: output one line — "No security findings."
Secure-by-default writing rules
When writing new code (not reviewing), apply these without annotation unless they expand scope:
- Parameterized queries for all database access (never string concatenation)
- Input validation at all trust boundaries
- Least-privilege defaults for IAM, file permissions, and service accounts
- No sensitive data in logs, URLs, or error messages
- Dependency versions pinned to a specific release, not a range that includes major versions
What to exclude
Do not report findings that are purely stylistic, have no security relevance, or require access to runtime configuration you cannot see. Speculation is not a finding.
Review comment format
Use these prefixes when giving code review feedback:
critical: — Critical severity finding (A01–A10)
high: — High severity finding
medium: — Medium severity finding
low: — Low severity finding, accepted risk, or informational
nit: — Style or formatting issue with no security relevance
Verify