Harden alpha grammar and auth-aware research

Introduce authoritative platform submission verification and boot-time reconciliation

Refine submitted seed mutation strategy, add fallback toggles, and introduce generation queue backoff

feat: implement QuantForge v2.5 payout optimizations: AST cloning filter, high-liquidity directives, and safe simulation timeout slot recycling

Resolve empty metrics envelopes by fetching alpha details endpoint and recover incomplete alphas automatically in main loop

Introduce ExpressionFactory grammar generator, store raw result payload in DB, and handle incomplete/timeout simulation statuses

Implement resilient retry logic, submission locks, rate-limit headers, and dynamic concurrency tracking

Implement wake_event synchronization to instantly resume recovery loop upon HTTP cookie updates

feat: Add Telegram API Proxy support to bypass Hugging Face firewall blocks

feat: Add keep-alive loop, Telegram auth recovery system, and fix main.py NameError crash

fix: rescue 71 stuck SIMULATING alphas on boot; lower Sharpe pre-filter 1.5->1.2 to unblock submissions

fix: send sanitized expression (1e-6->0.000001) to BRAIN payload; ban sci notation in prompt

fix: stop validation path from mutating expression column, add IntegrityError catch for both TRASHED code paths

fix: catch IntegrityError on duplicate expression to stop infinite 60s crash loop

fix: remove ts_skewness/ts_kurt, add ts_median/ts_zscore/ts_av_diff, block double group_neutralize, HTVR targeting

Execute Phase 11: Enforce strict vocabulary data fields and naked variable constraints in system prompt, and add sir2_ to validator prefixes

Execute Phase 9: Align OpenAI/NIM headers Content-Type casing, implement PV-safe fallback, and integrate Dynamic Pyramid Multiplier directives

Increase HTTP timeout to 90s, implement Nvidia key diversion for Zhipu provider, and restrict expression validator to base Price-Volume fields only

Phase 8: Free-tier optimization & fallback math fix - use ts_decay_linear in mock, silent key bypass, limit batch size to 5, 5s sleep, update Nemotron model name

Phase 7: Analytics dashboard overhaul - trash_reason tracking, enriched /api/stats, Semaphore(11), premium dark/light UI, LLM router model tracking, pause/resume endpoints

Expose MINER_RUNNING, MINER_TASK, and miner_loop in main.py for app.py dashboard compatibility

QuantForge v2.0 Phase 3: Autonomous Evaluation & Auto-Submission Pipeline and Infinite Orchestrator

Optimize Gemini key rotation, implement two-tier evaluation, and strict allowlist validator

fix: stagger simulation submissions and add exponential retry backoff to prevent CONCURRENT_SIMULATION_LIMIT_EXCEEDED (429) errors

fix: resolve 7 critical pipeline issues causing zero alpha submissions

Cap decay settings parameter at 512 to prevent API rejection

Migrate to event-driven immediate submission, increase daily cap to 20, and rename SP500 to TOPSP500

fetch and cache fields for all allowed universes (TOP3000, TOP2000, TOP1000, TOP500, TOP200, SP500)

allow multiple universes in settings and fallback mutations (TOP3000, TOP2000, TOP1000, TOP500, TOP200, SP500)

fix: sanitize sci notation for ALL queued alphas on boot + eliminate WARNING log spam from polling

fix: sanitize scientific notation format (1e-9) to standard decimal floats across generation and DB normalization phases, and update LLM prompts to prevent scientific notation output

Fix database normalization unique constraints boot crash, timer drift using client-side seconds, rate-limit retry logging noise, and add Details modal for Tested Archive

Fix variable hallucination: restrict local fallback fields to verified ones, add prompt warnings, normalize variables on startup, and prevent infinite resets on unknown variable errors

Implement evolutionary genetic loop, vector operator guardrails, automated orthogonalization, and smart retry backoffs

Implement expression/variable name validation, dynamic few-shot learning prompt injection, and fix pending checks sync bug

Realign submitted statuses, persist batch timer cooldown, and integrate categorized data fields library with Vector unit safeguards

Fix partial variable normalization and simulation result parsing when metrics are missing in poll response

Fix WorldQuant unknown variable 'ni' by mapping to 'income' and resolve AttributeError in simulation results parser

Fix database normalization unique constraint violation and prevent excessive logging

Replace invalid net_income/cash_flow variables with standard ni in DB and generator

Add dynamic concurrent simulations limit of 3 with UI control

Fix WorldQuant BRAIN API simulation payload settings and safe float Retry-After parsing

CRITICAL: Fix WQ API payload - nest settings under 'settings' key, reset bad-payload failures

Fix overnight zero-progress: per-model cooldown, expanded fallback, robust metric parsing, session re-auth, throttle fix

Implement startup recovery: reset stuck SIMULATING records back to GENERATED

Throttles Alpha Generation to at most once per 5 minutes to conserve free API limits

Initialize Space with QuantForge Miner code and 24/7 integration config