Public trust baseline

Honest claims users can verify.

TezzNative should be ambitious, but every public claim needs a matching doc, test, benchmark, or release artifact.

Current Position

TezzNative is strongest today for CLI tools, automation scripts, native utilities, small services, C interop experiments, and Windows/Linux x64 backend hardening.

The public message is: use TezzNative where Python feels slow and C feels painful.

Verified Claims

Claim	Evidence
Stable-core checks	docs/CONFORMANCE.md and tests/conformance/run.*
Windows/Linux native backend	docs/NATIVE_BACKEND.md plus run-native-smoke.* and run-native-reliability.*
Starter C ABI	docs/C_ABI.md and tests/conformance/run-abi.*
Benchmarks	docs/BENCHMARKS.md and benchmarks/
Release integrity	docs/RELEASE_ENGINEERING.md and release_manifest.json
Stability labels	docs/STABILITY.md, PLATFORM_SUPPORT.md, and STDLIB_INVENTORY.md

Boundaries

No broad replacement claim for every Python, C, Go, Rust, or Node.js use case.
GPU, NPU, LLM, kernel, OS, embedded, and GUI surfaces stay experimental until backend proof exists.
Public-network HTTP, TLS, DNS-backed sockets, and database production readiness remain beta or preview unless gated.
Performance claims require generated CSV and metadata from the benchmark harness.
ABI claims stay inside documented and tested starter layouts.