Goal
Submit one schema-valid deterministic baseline row so the leaderboard has more non-LLM evidence anchors instead of only maintainer-generated examples.
Suggested scope
Pick one simple deterministic policy and one documented scenario. Good options are always-hold, random, naive momentum, mean reversion, risk parity, or another classical baseline that does not call a live model provider.
Commands
python -m pip install -e ".[dev]"
python scripts/run_classical_baseline_matrix.py
tradearena validate-submission <path-to-your-row.json>
tradearena build-registry examples/benchmark_submissions \
--output docs/results/community_registry.md \
--csv-output docs/results/community_registry.csv \
--html-output docs/results/community_registry.html
Acceptance criteria
Please submit either an issue report or a PR that includes:
- deterministic policy name and scenario;
- manifest path;
- validation command output;
- reproducibility hash;
- registry diff or generated registry artifact path;
- evidence labels, especially deterministic-baseline, fully-auditable, and whether the row is external-submitted.
Keep the row free of provider prompts, API keys, broker credentials, private holdings, and live-order instructions.
Goal
Submit one schema-valid deterministic baseline row so the leaderboard has more non-LLM evidence anchors instead of only maintainer-generated examples.
Suggested scope
Pick one simple deterministic policy and one documented scenario. Good options are always-hold, random, naive momentum, mean reversion, risk parity, or another classical baseline that does not call a live model provider.
Commands
Acceptance criteria
Please submit either an issue report or a PR that includes:
Keep the row free of provider prompts, API keys, broker credentials, private holdings, and live-order instructions.