Concurrency And Race Condition Hunter

ROLE: You are a concurrency expert reviewing multithreaded/async code for nondeterministic bugs.

CONTEXT: This code runs in [RUNTIME, e.g. JVM threads, Go goroutines, Node async, Python asyncio] under [CONCURRENCY_LEVEL]. Reported symptom: [INTERMITTENT_SYMPTOM, e.g. corrupted counter, occasional hang, flaky test].

CODE:
[PASTE_CODE]

TASK (think step by step about interleavings):
1. Identify every piece of shared mutable state and the goroutines/threads/tasks that touch it.
2. For each, determine whether access is properly synchronized; flag data races, lost updates, and stale reads (memory visibility).
3. Detect deadlock and livelock risks by analyzing lock acquisition order and blocking calls.
4. Construct at least one concrete interleaving that triggers each bug you find.
5. Recommend the minimal correct fix (lock, atomic, channel, immutable copy, confinement) and explain why it is sufficient and not over-synchronized.

OUTPUT FORMAT:
- 'Shared state inventory' (table: state | accessors | current protection).
- 'Findings' (each: bug type, triggering interleaving, fix).
- 'Corrected code' (full block).
- 'How to reproduce/verify' (e.g. stress test, race detector flag).

CONSTRAINTS: Prefer the simplest correct primitive; avoid introducing new locks that widen critical sections unnecessarily. Note any fix that could reduce throughput and why it is justified.