Adriano d4fcb42fc5 feat(agents): hypothesis retry-with-error-feedback (max 1 retry) ...

HypothesisAgent.propose ora riprova una volta in caso di parse o
validation error: il prompt user del retry include l'output precedente
(troncato a 800 char) e il messaggio di errore, così l'LLM può
auto-correggersi. Configurabile via max_retries (default 1).

Cambia il modello dati di HypothesisProposal: completion (singolare)
diventa completions: list[CompletionResult] con n_attempts. L'orchestrator
itera su completions per registrare il costo di ogni chiamata LLM,
incluse le retry.

Phase 1 v4 mostrava 64% di parse failure recuperabili: il retry punta
a tagliare quel tasso senza inflazionare i token oltre 2x worst-case.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

2026-05-10 21:20:47 +02:00

..

__init__.py

feat(config): pydantic settings loader from .env

2026-05-09 18:55:04 +02:00

test_adversarial.py

refactor(protocol): swap S-expression grammar for strict JSON Schema

2026-05-10 21:17:26 +02:00

test_backtest_engine.py

feat(backtest): event-driven engine with 1-bar exec delay

2026-05-09 19:15:05 +02:00

test_backtest_orders.py

feat(backtest): Order/Position/Trade dataclasses with fees

2026-05-09 19:11:27 +02:00

test_cerbero_client.py

feat(cerbero): HTTP client with bearer + bot-tag + retry

2026-05-09 19:25:00 +02:00

test_cerbero_ohlcv.py

refactor(data): replace ccxt OHLCV loader with CerberoOHLCVLoader (deribit default)

2026-05-10 10:58:16 +02:00

test_cerbero_tools.py

feat(cerbero): tools wrapper for Phase 1 indicator subset

2026-05-09 19:27:34 +02:00

test_config.py

refactor(llm): route all tiers via OpenRouter, drop Anthropic SDK

2026-05-10 09:39:23 +02:00

test_cost_tracker.py

feat(llm): full multi-tier S/A/B/C/D with routing + pricing

2026-05-10 09:18:57 +02:00

test_falsification.py

refactor(protocol): swap S-expression grammar for strict JSON Schema

2026-05-10 21:17:26 +02:00

test_fitness.py

feat(ga): fitness v0 (DSR - dd_penalty * max_dd, kill on adversarial high)

2026-05-09 20:11:13 +02:00

test_ga_initial.py

feat(ga): initial population generator with cognitive style coverage

2026-05-09 20:22:04 +02:00

test_ga_loop.py

feat(ga): next_generation step (elitism + tournament + mutate/crossover)

2026-05-09 20:15:21 +02:00

test_ga_summary.py

feat(ga): generation summary stats (median/max/p90/entropy)

2026-05-09 20:20:12 +02:00

test_genome_crossover.py

feat(genome): uniform crossover for hypothesis genomes

2026-05-09 19:51:44 +02:00

test_genome_hypothesis.py

feat(llm): full multi-tier S/A/B/C/D with routing + pricing

2026-05-10 09:18:57 +02:00

test_genome_mutation.py

feat(genome): deterministic mutation operators (numeric + categorical)

2026-05-09 19:48:22 +02:00

test_hypothesis_agent.py

feat(agents): hypothesis retry-with-error-feedback (max 1 retry)

2026-05-10 21:20:47 +02:00

test_llm_client.py

refactor(llm): route all tiers via OpenRouter, drop Anthropic SDK

2026-05-10 09:39:23 +02:00

test_market_summary.py

feat(agents): market summary builder for hypothesis prompt

2026-05-09 20:24:23 +02:00

test_metrics_basic.py

feat(metrics): Sharpe + max drawdown + total return

2026-05-09 19:18:48 +02:00

test_metrics_dsr.py

feat(metrics): Deflated Sharpe Ratio (Bailey & Lopez de Prado)

2026-05-09 19:21:35 +02:00

test_protocol_compiler.py

refactor(protocol): swap S-expression grammar for strict JSON Schema

2026-05-10 21:17:26 +02:00

test_protocol_parser.py

refactor(protocol): swap S-expression grammar for strict JSON Schema

2026-05-10 21:17:26 +02:00

test_protocol_validator.py

refactor(protocol): swap S-expression grammar for strict JSON Schema

2026-05-10 21:17:26 +02:00

test_repository.py

feat(persistence): SQLite schema + repository for runs/genomes/evals/cost

2026-05-09 20:18:08 +02:00

test_selection.py

feat(ga): tournament selection + elitism

2026-05-09 20:12:51 +02:00

test_splits.py

feat(data): expanding walk-forward splits

2026-05-09 19:08:43 +02:00