aaditya-v-more/worldquant-orchestrator / skills/wqo-simulate
wqo-simulate
VerifiedBacktest alpha expressions and settings sweeps on WorldQuant BRAIN with WQO. Use for simulation requests, cached results and interpretation of backtest metrics.
WorldQuant BRAIN — simulation
Setup
Read setup and operational boundaries before running
commands. Use the bundled scripts/wqo.sh runner for every wqo command below;
it installs the CLI locally when missing. This skill works without other skills
or a source checkout.
One expression
wqo sim run --code "-ts_delta(ts_backfill(close, 60), 5)" \
--region USA --universe TOP3000 --delay 1 \
--neutralization SUBINDUSTRY --decay 6 --truncation 0.08
Returns JSON: alpha_id, url, sharpe, fitness, turnover, returns,
drawdown, margin, grade, checks_passed (e.g. "6/8").
grade is BRAIN's own verdict — INFERIOR, AVERAGE, GOOD, EXCELLENT,
SPECTACULAR — computed server-side. It is a hint about where to spend
attention, not an eligibility signal: checks_passed and wqo-analyze decide
what can actually be submitted.
Settings flags all have defaults; only pass what differs. Simulations can take a
few minutes — the command polls on the server's own Retry-After cadence and
blocks until done.
Out-of-sample tail
--test-period holds back the end of the backtest window:
wqo sim run --code "rank(close)" --test-period 1y
Accepts 1y, 6m, 1y6m or the ISO form P1Y6M. Default is none — the whole
window is in-sample. Reserving a tail shortens the window the reported Sharpe is
measured over, so numbers from different test periods are not comparable. Use it
to sanity-check one promising alpha, not as a standing default.
Batches and sweeps
Write a JSON list, then:
wqo sim batch --file ideas.json
[
{"code": "rank(close)", "label": "baseline"},
{"code": "zscore(volume)", "label": "vol", "settings": {"neutralization": "MARKET"}}
]
Plain strings work too. Per-entry settings merge over the command-line defaults,
so a settings sweep is the same expression repeated with different settings.
Concurrency is learned from the account, not assumed. A new account starts at 1 slot and the tool raises it only after clean runs. Don't try to force it higher.
Caching
Identical code + identical settings is not re-simulated; the prior result is
returned with "cached": true. Pass --no-cache to force a re-run.
wqo sim recent --limit 20
Shows the local ledger of past runs.
Writing the expression
Full rules in expression notes. The ones that cost a wasted slot to learn:
- No scientific notation.
1e-9fails withUnexpected character 'e'. Write0.000000001. - Backfill sparse fundamentals:
ts_backfill(field, 60). Without it the field is NaN most days, which silently concentrates the book into a handful of names and surfaces as an unrelated-lookingCONCENTRATED_WEIGHTfailure. - Check the operator exists —
wqo data operators. Availability depends on the account; an unavailable operator fails the simulation. - Groups are bare identifiers:
group_neutralize(rank(x), subindustry), not a quoted string. adv20is a real field inpv1. Use it for liquidity gating instead of recomputingts_mean(volume, 20).
Interpreting results
- A negative Sharpe may motivate testing a sign-reversed expression; re-simulate and re-check eligibility rather than assuming it is good.
checks_passedbelow the total means BRAIN flagged something; runwqo-analyzeorwqo gate <id>to see which check.- A failed simulation returns
"status": "ERROR"with the message. Review the status and sanitized diagnostic before considering another simulation.
Budget
Daily simulation budget defaults to 300 (WQO_SIM_BUDGET). Exit code 3 means it's
exhausted; stop and tell the user rather than raising the cap unilaterally.
Companion skill names above are optional guidance. If none are installed, use
the equivalent wqo command through this skill's bundled runner.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |