Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add DanielTomaro13/sportsdata-agents --skill backtest_designgit clone --depth 1 https://github.com/DanielTomaro13/sportsdata-agentsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/danieltomaro13/sportsdata-agents/backtest_design)<a href="https://agentmods.dev/skills/danieltomaro13/sportsdata-agents/backtest_design"><img src="https://agentmods.dev/badge/skills/danieltomaro13/sportsdata-agents/backtest_design/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/danieltomaro13/sportsdata-agents/backtest_design"><img src="https://agentmods.dev/badge/skills/danieltomaro13/sportsdata-agents/backtest_design.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00036 | $0.00509 |
| Opus 5 | $0.00018 | $0.00254 |
| Sonnet 5 | $0.00007 | $0.00102 |
| Haiku 4.5 | $0.00004 | $0.00051 |
Grade A, and why
backtest_design scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Backtest design — the ways backtests lie
A backtest is an argument, and most of them are flawed. Check these before quoting one.
Lookahead & point-in-time discipline
- The entry price must be one you could have GOT at prediction time. The platform's
run_backtestenforces this (entry = prevailing change-point atpredicted_at), so honestpredicted_attimestamps are part of the experiment — backdating a prediction to grab an early price is fabricating edge. - Features must be point-in-time too: no closing prices, no post-game stats, no season aggregates that include the predicted game.
Reading the report
- Lead with average CLV (quant_concepts): +CLV/−ROI = variance, the strategy is probably fine; −CLV/+ROI = luck, it probably isn't. ROI converges over hundreds of bets; CLV says something useful after dozens.
- Skips are findings:
no_price= warehouse coverage gap;no_result= settle the events;below_edge= the threshold did its job. A backtest silently built on 10% of predictions is a different experiment than claimed. - Sample size: under ~100 bets, confidence intervals on ROI span the whole conclusion. Quote the bet COUNT next to every headline number; call small samples anecdotes.
Multiple comparisons & survivorship
- Trying ten thresholds/feature-sets and reporting the best one is p-hacking: the winner's numbers are inflated by selection. Either pre-register one strategy or report ALL variants tried.
- Survivorship: backtesting only events that ended up with results/prices skews toward liquid, well-covered markets — say what fraction of the original universe the replay actually covered.
Iterating
- Change ONE thing per run (threshold, feature, model version) and keep the old report for the diff. The eval harness pins the golden replay; your experiments should be similarly reproducible — persist model versions, never overwrite.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 9d ago First seen · 45 lines · 36 tokens per session scan A c8569aa95797
backtest_design is a skill published in the GitHub repository DanielTomaro13/sportsdata-agents (6 stars, last pushed 6d ago), licensed MIT. It adds 36 tokens to every session and 509 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
comps-analysis
Build comparable-company valuation workbooks in Excel.
hyperliquid
Hyperliquid market data, account history, trade review.
stocks
Stock quotes, history, search, compare, crypto via Yahoo.
omni-budget
Configure spending limits, token quotas, and rate-limit policies per API key or globally. Inspect current consumption and enforce cost controls across providers.
polymarket
Query Polymarket: markets, prices, orderbooks, history.
industry-research
An industry-research workflow that maps how an industry works from suppliers to customers, then examines individual companies within it. The value-investing analysis looks at business quality, risks, and price.