Runtime Reliability

Runtime Reliability documents how 0x0 runtime behavior fails, recovers, and

reports problems.

Failure Semantics

Runtime failures must return stable diagnostics or process statuses. Streaming

IO, structured argv/env, dynamic dispatch, host effects, sockets, subprocesses,

and framework runtimes have negative tests for malformed inputs and denied

capabilities.

Recovery

Recovery evidence is split by surface:

• runtime memory and dynamic behavior: make runtime-depth-closure-check;
• host effects: make host-effects-closure-check;
• framework runtime: make framework-runtime-closure-check;
• distributed cluster: make distributed-cluster-check;
• physical hardware intake: make physical-hardware-evidence-check.

Operator Runbooks

Operational procedures live under docs/operator-runbooks/. Incident handling,

rollback, evidence review, physical opt-in, emulator, and virtual hardware

runbooks are part of release documentation.