AI / Daily Analyst

AI Agent

Daily result review, settings suggestions, dialogue, cron scheduling, and optional pre-deal approval/rejection.

Agent Enabled Pre-deal Enabled

Main model

nvidia_nim / qwen/qwen3-next-80b-a3b-instruct

NIM_API_KEY: configured

Same-provider fallbacks

nvidia/nemotron-mini-4b-instruct, meta/llama-3.3-70b-instruct

Same API key/base URL, different models.

External fallback

Disabled

Used after same-provider fallbacks fail.

Cron

0 9 * * *

Last run: n/a

Regime adaptation

Enabled

Period: 1h; auto-apply: off

Telegram AI bot

Configured

Thread: 85902

Incident Cards

Compact diagnosis for the common AI-agent and setup-flow failure classes.

LLM connectivity

WARNING

9 recent provider issue(s). Latest: nvidia_nim meta/llama-3.3-70b-instruct is cooling down until 2026-07-07T11:05:52.233341+00:00 after timeout: ReadTimeout: provider request timed out after 45s

Use Test main LLM, check model name/key/base URL, then review Logs.

Setup flow

ONLINE

10 setup record(s) in the last 10 hours.

Use shadow score and daily report before changing enforcement.

Tick delivery

NO DATA

Telegram execution alerts are on, but tick velocity alerts are disabled.

Enable tick alerts only when you want Telegram velocity notifications.

Daily cron

NO DATA

Last status: idle.

Run daily check manually after changing models or cron.

AI Playbooks

One-click AI action drafts. Every playbook creates a pending approval item before it can affect runtime state.

AI Action Queue

Telegram, chat, and playbooks queue validated actions here. Approval records rollback and audit metadata.

No queued AI actions yet.

Apply Suggested Settings

The AI can propose a JSON patch for manual approval. Explicit UI or Telegram chat requests can also apply validated StrategyRuntimeConfig keys.

NO DATA llmGenerated: 07.07.2026 17:59:19nvidia_nim / meta/llama-3.1-8b-instruct

Market regime is bull_trend with low_liquidity_chop dominant observed regime; prefer trend profile behavior without removing risk gates.

39 active candidates vs 41 monitor candidates.
Dominant observed regime: low_liquidity_chop.
Average near-miss volume_z: 2.46.

Ignored keys

changed_key: unknown_strategy_key

AI Shadow Score

Compares AI approve/reject decisions with virtual trade outcomes before enforcing live rejections.

No AI verdict outcomes yet. Enable pre-deal analysis in shadow mode, keep enforcement off, then let virtual trades resolve.

Settings

API keys and the Telegram bot token stay in env; runtime choices are also mirrored to .env.

AI Agent enabled Enables daily analysis, Telegram replies, and LLM-backed regime adaptation. Tool access Allows the agent to call approved bot tools for settings, news, calendar, regime, correlation, and backtest context. Pre-deal analysis Adds an AI verdict to each new setup. Enforce AI rejections Rejected setups become suppressed audit records. Market regime adaptation Runs periodic regime classification and proposes StrategyRuntimeConfig patches. Regime adaptation period Controls how often the AI agent classifies regime and proposes/applies tuning. Auto-apply regime patch Applies validated regime patches automatically. Leave off for manual review. Telegram bot token

Saved to AI_AGENT_BOT_TOKEN in .env and applied to the running process. Telegram topic/thread id Leave empty to answer mentions in every topic from ALLOWED_CHAT_IDS; set a topic id to restrict replies and scheduled suggestions. Main provider Main base URL Main API key

Saved to NIM_API_KEY in .env and applied to the running process. Main model

You can type manually or choose a preset/model returned by the provider. 2nd fallback model

You can type manually or choose a preset/model returned by the provider. 3rd fallback model

You can type manually or choose a preset/model returned by the provider. External fallback provider External fallback base URL External fallback API key

Saved to provider default env in .env and applied to the running process. External fallback model

You can type manually or choose a preset/model returned by the provider. Cron schedule Five fields: minute hour day month weekday. Max tokens Temperature Request timeout Pre-deal timeout

ONLINE Main key env: NIM_API_KEY NO DATA Fallback key env: n/a ONLINE Telegram bot env: AI_AGENT_BOT_TOKEN

Slot	Status	Provider	Model	Cooldown until	Last error
Main	NO DATA timeout	nvidia_nim	qwen/qwen3-next-80b-a3b-instruct	2026-07-07T11:05:06.139464+00:00	ReadTimeout: provider request timed out after 45s

Useful skills to add next

Post-mortem clustering by reason: liquidity, late breakout, trend mismatch, cooldown, ML/AI rejection.
Risk-mode recommendation: normal, defensive, or pause suggestions based on daily drawdown and winrate.
Settings diff proposal: generate a JSON patch, then require manual approval before applying it.
Telegram summary delivery after daily run with links back to STAT and SETUP pages.
Shadow-score tracking: compare AI approved/rejected outcomes before enforcing it live.

Daily Report

Schedule state: idle

No AI daily report yet. Run one manually or enable the cron schedule.

Trade Journal

AI post-mortems for closed virtual trades: conditions, trigger, outcome, and lesson learned.

JSON

No AI trade journals yet. They will appear after virtual trades close.

Dialogue

Ask the agent about today, settings, risk gates, or why recent deals were filtered.

No dialogue yet.

Memory

Stored in Redis and included in future prompts together with recent dialogue.

Interactions0

Updatedn/a

No memory yet.

Remembered facts

No explicit preferences remembered yet.

Logs

Provider calls, missing keys, scheduler actions, memory updates, and model loading.

Export logs

Time	Event	Payload
07.07.2026 18:00:55	market_regime_adaptation_scheduler_finished	{"auto_apply": false, "changes": 0, "regime": "bull_trend", "status": "empty"}
07.07.2026 18:00:55	market_regime_adaptation_completed	{"applied": false, "auto_apply": false, "changes": 0, "regime": "bull_trend", "source": "llm"}
07.07.2026 18:00:55	provider_same_provider_fallback_completed	{"failed_model": "qwen/qwen3-next-80b-a3b-instruct", "fallback_model": "meta/llama-3.1-8b-instruct", "provider": "nvidia_nim"}
07.07.2026 18:00:55	provider_request_completed	{"model": "meta/llama-3.1-8b-instruct", "provider": "nvidia_nim", "response_chars": 361}
07.07.2026 18:00:53	provider_request_started	{"model": "meta/llama-3.1-8b-instruct", "provider": "nvidia_nim", "url": "https://integrate.api.nvidia.com/v1/chat/completions"}
07.07.2026 18:00:53	provider_same_provider_fallback_started	{"failed_model": "qwen/qwen3-next-80b-a3b-instruct", "fallback_model": "meta/llama-3.1-8b-instruct", "provider": "nvidia_nim"}
07.07.2026 18:00:53	provider_same_provider_fallback_failed	{"error": "nvidia_nim meta/llama-3.3-70b-instruct is cooling down until 2026-07-07T11:05:52.233341+00:00 after timeout: ReadTimeout: provider request timed out after 45s", "fallback_model": "meta/llama-3.3-70b-instruct", "provider": "nvidia_nim"}
07.07.2026 18:00:53	provider_cooldown_skip	{"cooldown_until": "2026-07-07T11:05:52.233341+00:00", "model": "meta/llama-3.3-70b-instruct", "provider": "nvidia_nim"}
07.07.2026 18:00:53	provider_same_provider_fallback_started	{"failed_model": "qwen/qwen3-next-80b-a3b-instruct", "fallback_model": "meta/llama-3.3-70b-instruct", "provider": "nvidia_nim"}
07.07.2026 18:00:53	provider_same_provider_fallback_failed	{"error": "nvidia_nim nvidia/nemotron-mini-4b-instruct is cooling down until 2026-07-07T11:05:06.833170+00:00 after http_error: Provider HTTP 500: {\"error\":\"'bool' object has no attribute 'get'\"}", "fallback_model": "nvidia/nemotron-mini-4b-instruct", "provider": "nvidia_nim"}
07.07.2026 18:00:53	provider_cooldown_skip	{"cooldown_until": "2026-07-07T11:05:06.833170+00:00", "model": "nvidia/nemotron-mini-4b-instruct", "provider": "nvidia_nim"}
07.07.2026 18:00:53	provider_same_provider_fallback_started	{"failed_model": "qwen/qwen3-next-80b-a3b-instruct", "fallback_model": "nvidia/nemotron-mini-4b-instruct", "provider": "nvidia_nim"}
07.07.2026 18:00:53	provider_main_failed	{"error": "nvidia_nim qwen/qwen3-next-80b-a3b-instruct is cooling down until 2026-07-07T11:05:06.139464+00:00 after timeout: ReadTimeout: provider request timed out after 45s", "model": "qwen/qwen3-next-80b-a3b-instruct", "provider": "nvidia_nim"}
07.07.2026 18:00:53	provider_cooldown_skip	{"cooldown_until": "2026-07-07T11:05:06.139464+00:00", "model": "qwen/qwen3-next-80b-a3b-instruct", "provider": "nvidia_nim"}
07.07.2026 18:00:53	tools_completed	{"tool_calls": ["apply_strategy_settings"]}
07.07.2026 18:00:53	tool_executed	{"elapsed_ms": 0, "status": "error", "tool": "apply_strategy_settings"}
07.07.2026 18:00:53	provider_same_provider_fallback_completed	{"failed_model": "qwen/qwen3-next-80b-a3b-instruct", "fallback_model": "meta/llama-3.1-8b-instruct", "provider": "nvidia_nim"}
07.07.2026 18:00:53	provider_request_completed	{"model": "meta/llama-3.1-8b-instruct", "provider": "nvidia_nim", "response_chars": 6}
07.07.2026 18:00:52	provider_request_started	{"model": "meta/llama-3.1-8b-instruct", "provider": "nvidia_nim", "url": "https://integrate.api.nvidia.com/v1/chat/completions"}
07.07.2026 18:00:52	provider_same_provider_fallback_started	{"failed_model": "qwen/qwen3-next-80b-a3b-instruct", "fallback_model": "meta/llama-3.1-8b-instruct", "provider": "nvidia_nim"}
07.07.2026 18:00:52	provider_same_provider_fallback_failed	{"error": "ReadTimeout: provider request timed out after 45s", "fallback_model": "meta/llama-3.3-70b-instruct", "provider": "nvidia_nim"}
07.07.2026 18:00:52	provider_transport_error	{"error": "ReadTimeout: provider request timed out after 45s", "model": "meta/llama-3.3-70b-instruct", "provider": "nvidia_nim"}
07.07.2026 18:00:06	provider_request_started	{"model": "meta/llama-3.3-70b-instruct", "provider": "nvidia_nim", "url": "https://integrate.api.nvidia.com/v1/chat/completions"}
07.07.2026 18:00:06	provider_same_provider_fallback_started	{"failed_model": "qwen/qwen3-next-80b-a3b-instruct", "fallback_model": "meta/llama-3.3-70b-instruct", "provider": "nvidia_nim"}
07.07.2026 18:00:06	provider_same_provider_fallback_failed	{"error": "Provider HTTP 500: {\"error\":\"'bool' object has no attribute 'get'\"}", "fallback_model": "nvidia/nemotron-mini-4b-instruct", "provider": "nvidia_nim"}
07.07.2026 18:00:06	provider_http_error	{"detail": "Provider HTTP 500: {\"error\":\"'bool' object has no attribute 'get'\"}", "model": "nvidia/nemotron-mini-4b-instruct", "provider": "nvidia_nim", "status_code": 500}
07.07.2026 18:00:06	provider_request_started	{"model": "nvidia/nemotron-mini-4b-instruct", "provider": "nvidia_nim", "url": "https://integrate.api.nvidia.com/v1/chat/completions"}
07.07.2026 18:00:06	provider_same_provider_fallback_started	{"failed_model": "qwen/qwen3-next-80b-a3b-instruct", "fallback_model": "nvidia/nemotron-mini-4b-instruct", "provider": "nvidia_nim"}
07.07.2026 18:00:06	provider_main_failed	{"error": "ReadTimeout: provider request timed out after 45s", "model": "qwen/qwen3-next-80b-a3b-instruct", "provider": "nvidia_nim"}
07.07.2026 18:00:06	provider_transport_error	{"error": "ReadTimeout: provider request timed out after 45s", "model": "qwen/qwen3-next-80b-a3b-instruct", "provider": "nvidia_nim"}
07.07.2026 17:59:20	provider_request_started	{"model": "qwen/qwen3-next-80b-a3b-instruct", "provider": "nvidia_nim", "url": "https://integrate.api.nvidia.com/v1/chat/completions"}
07.07.2026 16:59:42	market_regime_adaptation_scheduler_finished	{"auto_apply": false, "changes": 0, "regime": "sideways_chop", "status": "empty"}
07.07.2026 16:59:42	market_regime_adaptation_completed	{"applied": false, "auto_apply": false, "changes": 0, "regime": "sideways_chop", "source": "llm"}
07.07.2026 16:59:42	provider_request_completed	{"model": "qwen/qwen3-next-80b-a3b-instruct", "provider": "nvidia_nim", "response_chars": 794}
07.07.2026 16:59:25	provider_request_started	{"model": "qwen/qwen3-next-80b-a3b-instruct", "provider": "nvidia_nim", "url": "https://integrate.api.nvidia.com/v1/chat/completions"}
07.07.2026 16:59:25	tools_completed	{"tool_calls": ["get_market_regime_context"]}
07.07.2026 16:59:25	tool_executed	{"elapsed_ms": 82, "status": "ok", "tool": "get_market_regime_context"}
07.07.2026 16:59:25	provider_request_completed	{"model": "qwen/qwen3-next-80b-a3b-instruct", "provider": "nvidia_nim", "response_chars": 0}
07.07.2026 16:59:15	provider_request_started	{"model": "qwen/qwen3-next-80b-a3b-instruct", "provider": "nvidia_nim", "url": "https://integrate.api.nvidia.com/v1/chat/completions"}
07.07.2026 15:59:08	market_regime_adaptation_scheduler_finished	{"auto_apply": false, "changes": 0, "regime": "dead_market", "status": "empty"}
07.07.2026 15:59:08	market_regime_adaptation_completed	{"applied": false, "auto_apply": false, "changes": 0, "regime": "dead_market", "source": "llm"}
07.07.2026 15:59:08	provider_request_completed	{"model": "qwen/qwen3-next-80b-a3b-instruct", "provider": "nvidia_nim", "response_chars": 510}
07.07.2026 15:58:59	telegram_bot_connected	{"allowed_chats": 1, "bot_thread_id": 85902, "bot_username": "luma_sma_bot"}
07.07.2026 15:58:57	provider_request_started	{"model": "qwen/qwen3-next-80b-a3b-instruct", "provider": "nvidia_nim", "url": "https://integrate.api.nvidia.com/v1/chat/completions"}
07.07.2026 15:58:57	telegram_alive_sent	{"bot_thread_id": 85902, "chats": 1}
07.07.2026 15:58:55	settings_saved	{"api_key_env": "NIM_API_KEY", "bot_thread_id": 85902, "bot_token_configured": true, "daily_schedule": "0 9 * * *", "enabled": true, "fallback_api_key_env": "", "fallback_model": "", "fallback_provider": "", "market_regime_adaptation_auto_apply": false, "market_regime_adaptation_enabled": true, "market_regime_adaptation_period": "1h", "model": "qwen/qwen3-next-80b-a3b-instruct", "predeal_enabled": true, "predeal_enforce_rejections": false, "provider": "nvidia_nim", "same_provider_fallback_model_1": "nvidia/nemotron-mini-4b-instruct", "same_provider_fallback_model_2": "meta/llama-3.3-70b-instruct", "tools_enabled": true}
07.07.2026 15:58:55	api_key_saved	{"api_key_env": "NIM_API_KEY", "env_file": ".env", "provider": "nvidia_nim"}
07.07.2026 15:58:55	telegram_bot_token_saved	{"env": "AI_AGENT_BOT_TOKEN", "env_file": ".env"}