strategy-passport

Interpret strategy passport records and rigor gate verdicts without overstating statistical claims.

17|3|Updated May 12, 2026
One-click install
npx skills add https://github.com/aprin-labs/archimedes --skill strategy-passport-aprin-labs
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: strategy-passport
Source: https://github.com/aprin-labs/archimedes/tree/main/skills/strategy-passport
Command: npx skills add https://github.com/aprin-labs/archimedes --skill strategy-passport-aprin-labs

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Summarizing a quantitative trading strategy's backtest and rigor-gate results is easy to get wrong: a status="live" field sounds like a rigor endorsement, a bare passes_rigor_gate boolean hides pending states, and phrases like "statistically proven" or "corrects for multiple testing" contradict what the code actually does. This Skill maps every field of the strategy_passports table to its true meaning and enforces honest framing. ## Core Features & Use Cases - Field-by-field reference: Documents every column of StrategyPassportRecord (identity, lifecycle, ownership, backtest metrics, rigor gate results) with source line numbers. - Gate semantics: Explains what passes_rigor_gate and the tri-state rigor_gate_status mean, including that the badge is always graded at the strictest level and that curated strategies are graded at num_trials=1. - Claim integrity rules: Lists five forbidden phrasings (e.g., "statistically proven", "95% confidence") plus five pitch anti-claims that must never appear in strategy summaries. - Use Case: When a user asks "does this strategy pass the rigor gate?", read the passport via GET /api/strategies/passports/{id}, report DSR, PBO, and out-of-sample Sharpe with their exact scope, and explicitly separate status from the gate verdict. ## Quick Start Read the passport for strategy ID X and summarize its rigor gate results honestly, stating whether status is live and whether the gate passed as separate facts.

Frequently Asked Questions about strategy-passport

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I check if a strategy passes the rigor gate?

Fetch the passport via GET /api/strategies/passports/{id} and read rigor_gate_status, which returns pending, pass, or fail. Prefer it over the bare passes_rigor_gate boolean, since a false value cannot distinguish a failed strategy from one never backtested.

Does status=live mean a strategy passed the rigor gate?

No. Status live means the strategy is active in at least one portfolio or was hand-declared live by a curator; it is a usage fact, not a rigor verdict. A strategy can be status=live and rigor_gate_status=fail simultaneously.

What does passes_rigor_gate actually measure?

It is a stored boolean reflecting whether the strategy passed DSR, PBO, out-of-sample, and look-ahead checks at the strictest level when computed. It reduces the false-positive rate of backtest selection but says nothing about future returns.

Why is a curated strategy's DSR not deflated for multiple testing?

Curated strategies are graded at num_trials=1, meaning DSR runs undeflated against only their own return series. Only generated strategies are deflated, and only against their own candidate pool, so never claim the gate corrects for multiple testing.

What claims are forbidden when summarizing a strategy passport?

Never say the gate corrects for multiple testing, corrects selection bias, is statistically proven, uses 95% confidence (the live gate is 90% one-sided), or makes a strategy right. Also avoid predicted alpha, future-return guarantees, and production-readiness claims.

Which field determines strategy ownership for access checks?

owner_user_id is the canonical ownership field gating visibility and access after the Better Auth cutover. owner_wallet records wallet provenance only and must not be used for ownership checks.