llmquant-prediction-markets — Claude Code skill for prediction markets
A router skill for prediction-market work: event odds, settlement criteria, cross-venue arbitrage and options-implied probabilities.
npx skills add LLMQuant/skills --skill llmquant-prediction-markets -a claude-codeWhat llmquant-prediction-markets does
The skill breaks a request into its parts: event, venue, contract, settlement rule, deadline, related assets and the output you want. Then it picks one of three workflows and opens only that one — an event-probability brief, a cross-venue arbitrage check, or a comparison of market odds against options-implied and asset-implied pricing.
Market data comes from LLMQuant Data: contracts, odds, order-book depth, volume, fees, close dates, plus news, macro releases and related asset prices. Reports keep contract timestamps, settlement criteria, liquidity and market windows. If market data or settlement rules are missing, the skill will not infer fair probability or call something arbitrage — that is your cue to check what the conclusion actually rests on. Odds you type in yourself get labelled as user supplied.
Install runs through npx skills add, and the skill file lands at ~/.claude/skills/llmquant-prediction-markets/SKILL.md.
What it helps with
- build an event-probability brief from market odds
- check arbitrage across venues and contracts
- compare prediction-market odds with options-implied pricing
- pin down settlement criteria and the contract deadline
- separate user-supplied odds from market data
Example prompts for the agent
- Estimate the probability of this event from the contracts on that venue and write a brief
- Check whether there is arbitrage between the two venues on this event
- Compare the prediction-market odds with what the options are pricing in
How to install llmquant-prediction-markets
1Skills CLI (Vercel)
npx skills add LLMQuant/skills --skill llmquant-prediction-markets -a claude-code2Manually: put the files in place
~/.claude/skills/llmquant-prediction-markets/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/llmquant-prediction-markets/
1Skills CLI (Vercel)
npx skills add LLMQuant/skills --skill llmquant-prediction-markets -a codex2Manually: put the files in place
.agents/skills/llmquant-prediction-markets/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add LLMQuant/skills --skill llmquant-prediction-markets -a cursor2Manually: put the files in place
.cursor/skills/llmquant-prediction-markets/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install LLMQuant/skills llmquant-prediction-markets2Manually: put the files in place
.github/skills/llmquant-prediction-markets/SKILL.md1Skills CLI (Vercel)
npx skills add LLMQuant/skills --skill llmquant-prediction-markets -a gemini-cli2Platform CLI
gemini skills install https://github.com/LLMQuant/skills --path skills/llmquant-prediction-marketsManually: put the files in place
.devin/skills/llmquant-prediction-markets/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add LLMQuant/skills --skill llmquant-prediction-markets -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about llmquant-prediction-markets
How do I install llmquant-prediction-markets in Claude?
Run npx skills add LLMQuant/skills --skill llmquant-prediction-markets -a claude-code. The skill ends up at ~/.claude/skills/llmquant-prediction-markets/SKILL.md.
What does llmquant-prediction-markets do?
It routes prediction-market requests: it identifies the event, venue and contract, then runs one of three workflows — a probability brief, an arbitrage check, or a comparison against options pricing.
What does the skill need to work?
LLMQuant Data: contracts, odds, liquidity, options, macro and news. Without market data or settlement rules it will not infer arbitrage or a fair probability.
Source text
SKILL.md · MIT
--- name: llmquant-prediction-markets description: Router skill for LLMQuant prediction-market workflows. Use when the user needs event odds, settlement criteria, probability gaps, cross-market pricing, or prediction-market arbitrage review. input_data_source: LLMQuant Data category: prediction-markets --- # LLMQuant Prediction Markets This category routes event-probability workflows for prediction markets, options-implied probabilities, and tradable event risk. ## Routing Rules 1. Identify the event, venue, contract, settlement rule, deadline, related assets, and requested output. 2. Select the closest workflow below. 3. Open only that workflow and any referenced local resources. 4. Use LLMQuant Data for prediction-market contracts, prices, liquidity, options, macro, news, and related asset prices. 5. Report contract timestamps, settlement criteria, liquidity, fees, market windows, and missing inputs. ## Workflow Index | User intent | Workflow | |---|---| | Produce an event probability research brief from market odds and evidence. | [`workflows/event-probability-brief.md`](workflows/event-probability-brief.md) | | Check prediction-market cross-venue or contract-level arbitrage conditions. | [`workflows/prediction-market-arb-watch.md`](workflows/prediction-market-arb-watch.md) | | Compare prediction-market odds with options-implied or asset-implied event pricing. | [`workflows/probability-vs-options-pricing.md`](workflows/probability-vs-options-pricing.md) | ## LLMQuant Data Contract Prefer LLMQuant Data when available. The workflows may need these data capabilities: - Retrieve event contracts, settlement criteria, market odds, order-book depth, volume, fees, and close dates. - Retrieve related news, macro releases, asset prices, and issuer or sector context. - Retrieve options-implied probabilities, volatility, skew, and event-window pricing when available. - Compare venues, contracts, and outcome sets while preserving timestamp and settlement-rule differences. Fallback: - If market data or settlement rules are unavailable, do not infer arbitrage or fair probability. - If only user-provided odds are available, label the evidence as user supplied.
© LLMQuant · MIT · LLMQuant/skills