trading-agent/orchestration/run_nightly.py

"""ENTRYPOINT #1 — the nightly research run.

Brings up oMLX, then runs the 3 personas strictly sequentially (one model, time-sliced),
each under a rollover wall-clock budget bounded by the hard stop. Every agent's output is
checkpointed to state/output/<trade_date>/. Run with --compressed for a fast end-to-end
test (short budgets).

    uv run python -m orchestration.run_nightly [--compressed]
"""

from __future__ import annotations

import sys
import time
from datetime import datetime, timezone
from pathlib import Path

sys.path.insert(0, str(Path(__file__).resolve().parent.parent))

from loguru import logger  # noqa: E402

from agents.graph import run_agent  # noqa: E402
from agents.schemas import RunManifest  # noqa: E402
from config.agent_registry import load_personas  # noqa: E402
from config.settings import get_settings  # noqa: E402
from llm import server  # noqa: E402
from llm.client import resolve_model_id  # noqa: E402
from orchestration.budget import (  # noqa: E402
    local_hard_stop_ts,
    next_deadline,
    next_market_trade_date,
)
from orchestration import committee as committee_mod  # noqa: E402
from orchestration.checkpoint import Checkpointer  # noqa: E402
from orchestration.regime import market_regime  # noqa: E402
from reporting import report_writer  # noqa: E402
from reporting.journal import lessons_for  # noqa: E402


def main(compressed: bool = False) -> int:
    s = get_settings()
    s.ensure_dirs()
    trade_date = next_market_trade_date(s.market_tz)
    logger.add(s.log_dir / f"nightly_{trade_date}.log", rotation="10 MB", retention="14 days")
    logger.info("=== nightly run trade_date={} profile={} compressed={} ===",
                trade_date, s.model_profile, compressed)

    if not server.ensure_ready():
        logger.error("oMLX not ready; aborting.")
        return 1

    # preflight: confirm internet before committing to a long run (warn but continue)
    try:
        from tools.web import web_search

        probe = web_search.invoke({"query": "stock market news today"})
        logger.info("internet preflight: {}", "OK" if "http" in probe else "no results (degraded)")
    except Exception as e:
        logger.warning("internet preflight failed (continuing): {}", e)

    model_name = resolve_model_id()
    run_id = datetime.now(timezone.utc).strftime("%Y%m%dT%H%M%SZ")
    cp = Checkpointer(s.output_dir, trade_date)
    manifest = RunManifest(
        run_id=run_id,
        trade_date=trade_date,
        model=model_name,
        started_at=datetime.now(timezone.utc),
        status="running",
    )
    cp.write_manifest(manifest)

    hard_stop = local_hard_stop_ts(s.hard_stop, s.local_tz)
    step_budget = 12 if compressed else s.per_agent_step_budget
    min_steps = 4 if compressed else s.per_agent_min_steps

    # dashboard-managed registry: built-ins + overrides + custom agents, disabled excluded
    personas = load_personas()
    logger.info("agents tonight: {}", ", ".join(p.id for p in personas))

    # shared ground truth for all agents tonight (computed, deterministic)
    regime = market_regime()
    if regime:
        logger.info("regime: {}", regime[:140])

    outputs = []
    records: dict[str, str] = {}
    for i, persona in enumerate(personas):
        # per-agent budgets (dashboard-editable), falling back to the global settings
        if compressed:
            per_agent_cap_s, min_seconds = 5 * 60, 60
        else:
            per_agent_cap_s = (persona.time_budget_min or s.per_agent_wallclock_min) * 60
            min_seconds = (persona.min_research_min or s.per_agent_min_minutes) * 60
        now = time.time()
        if now >= hard_stop:
            logger.warning("hard stop reached; skipping {} remaining agents", len(personas) - i)
            break
        deadline = next_deadline(now, hard_stop, len(personas) - i, per_agent_cap_s)
        logger.info("--- {} until {} ---", persona.id,
                    datetime.fromtimestamp(deadline).strftime("%H:%M:%S"))
        if not server.server_up():  # recover if a prior agent left oMLX unresponsive
            logger.warning("oMLX unreachable before {}; recovering...", persona.id)
            server.ensure_ready(warm=False)
        lessons = lessons_for(persona.id)  # self-learning: past picks marked to market
        records[persona.id] = lessons
        blurb = "\n\n".join(x for x in (regime, lessons) if x)
        out = run_agent(persona, deadline_ts=deadline, step_budget=step_budget,
                        min_steps=min_steps, min_seconds=min_seconds,
                        context_blurb=blurb, checkpointer=cp, model_name=model_name)
        cp.write(out)
        outputs.append(out)
        manifest.agents.append(persona.id)
        try:
            report_writer.write_agent_report(out, trade_date)  # analyst research note
        except Exception as e:
            logger.warning("agent report failed for {}: {}", persona.id, e)

    # Investment Committee: risk-officer pass over the whole book (verdicts enforce at open)
    try:
        committee_mod.run_committee(outputs, trade_date, regime, records)
    except Exception as e:
        logger.warning("committee failed (fail-open): {}", e)

    # combined daily report (execution results appended in the morning by run_market_open)
    try:
        daily = report_writer.write_daily_report(trade_date, outputs)
        report_writer.to_pdf(daily)
    except Exception as e:
        logger.warning("daily report failed: {}", e)

    manifest.status = "complete"
    manifest.ended_at = datetime.now(timezone.utc)
    cp.write_manifest(manifest)
    logger.info("=== nightly run complete; output in {} ===", cp.dir)
    return 0


if __name__ == "__main__":
    raise SystemExit(main(compressed="--compressed" in sys.argv))