The scanner is not
- a full security audit
- a compliance review
- a guaranteed vulnerability scanner
- a guaranteed modernization planner
- an autonomous production rewrite engine
- a substitute for developer review
- a substitute for legal, compliance, or security review
Current private scan limitations
- static indicators only unless AI deep scan is explicitly enabled
- no guarantee that all risks are detected
- no guarantee that all secrets are found
- no guarantee that code is safe to deploy
- no guarantee that a modernization path is complete
Memory-Safe Modernization Roadmap
Modernize Agent AI currently focuses on static legacy-code risk scanning, evidence-backed reports, callback/domain review, archive/package intake, and modernization readiness planning.
Future roadmap areas may include:
- C/C++ to Rust readiness assessment
- memory-safety risk identification
- behavior-preservation planning
- test/readiness gap detection
- containment-first recommendations
- security profile and sandboxing considerations
- migration planning for high-risk legacy components
Limitations:
- Modernize Agent AI does not currently perform automatic C-to-Rust conversion.
- Modernize Agent AI does not currently prove behavior equivalence.
- Modernize Agent AI does not currently perform AppArmor profile generation or enforcement.
- Modernize Agent AI does not currently certify generated code as safe.
- Any generated or transformed code must be treated as untrusted until validated.
Evidence doctrine
A recode is not successful because it compiles. A recode is successful only when behavior, risk, dependencies, security posture, and modernization goals are validated with evidence.
Safe usage
Use scan results as planning input. Review, test, and validate before making production changes.