feat: track local-path usage and enrich analytics dashboard

feat: add usage tracking with live throughput rates and analytics dashboard

Preserve empty reasoning state across native tool turns (#1339)

Restore Claude Code client controls for OpenAI Responses (#1311)

Keep tool names portable across OpenAI transports (#1309)

Add ChatGPT OAuth as an OpenAI Codex provider (#1279)

Close interrupted tool turns before new user input (#1271)

Recover MiniMax-M3 native tool calls on NVIDIA NIM (#1246)

Recover Claude sessions from provider context overflow (#1215)

Coordinate provider recovery under concurrent load (#1205)

Report output-limit truncation as incomplete Responses (#1180)

Honor reasoning tiers on numeric-budget providers (#1150)

Keep inline system reminders cache-stable across providers (#1154)

Make reasoning policy provider-neutral and client-aware (#1148)

Add LOG_LEVEL env var to control log verbosity (#1142)

Preserve mid-conversation system messages through provider conversion (#1125)

Add Ollama Cloud as a first-class provider (#1106)

Fix Auto Mode classifier responses when stream is omitted (#1098)

feat: add vision support for OpenAI chat conversion (#1061)

Consolidate OpenAI-compatible provider adapters (#1069)

Remove the native Anthropic provider transport (#1067)

Make response stream lifetimes explicit (#1060)

Make provider backoff admission monotonic (#1053)

Use installed metadata as the single package version source (#1051)

Canonicalize provider failure and retry ownership (#1046)

Introduce a typed application boundary for provider execution (#1045)

Move protocol models to their protocol owners (#1044)

Establish single-owner runtime with stream-safe provider hot swaps (#1036)

Restore protocol-correct provider errors without client retry loops (#1033)

Move runtime packages under src namespace (#1029)