Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Research prediction-market events, venues, underliers, liquidity, and news context for Itô basket workflows. Use for read-only market intelligence, API-gated Itô exploration, and source-grounded prediction-market briefings without investment advice or live trading.
.claude/skills/ito-market-intelligence/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-13 | ✗→✓ | ▲ Improved | 85% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -27% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 254% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 85% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 15% | 0% |
Use this skill when a user wants prediction-market context, event discovery, venue comparison, basket theme exploration, or an Itô API-backed market brief.
Use public sources by default. Any Itô-backed data call requires the user to explicitly request Itô data and requires a scoped ITO_API_KEY. Never print, persist, or ask the user to paste a key into chat.
not as truth by themselves.
event that is absent from a cited response or source.
URL, and source-provided update time when available. Call data stale or unknown rather than silently treating it as current.
Cite the exact source URL next to each material claim and distinguish the publication/update time from the retrieval time.
bash node scripts/ito-market-intelligence.js --json search-markets --platform all --limit 25
The client reads ITO_API_KEY from the environment, sends it only to the configured Itô HTTPS origin, never logs it, and permits only documented GET endpoints. Do not run it merely because a key exists.
differences across venues.
Do not run ito login, ecc ito login, or open a browser for this skill: those credentials are not a documented substitute for a baskets:read or markets:read API key. Return control to the originating agent after stating the missing scope and operator-driven access requirement.
AUTH_MISSING, request a scoped key through the user's established Itôaccess channel without collecting it in chat. On AUTH_REJECTED, say the key may be expired, revoked, or missing the required read scope.
RATE_LIMITED, respect retry_after_seconds; do not loop automatically.On TIMEOUT or UPSTREAM_ERROR, preserve prior cited facts, label the live snapshot unavailable, and offer a bounded retry. Never replace failed live data with invented values.
ITO_MARKET_API_URL may override the API origin for deterministic localtests. In normal use keep the default https://itomarkets.com/api/v1.
deep-research or exa-search for source discovery.x-api for public social signal discovery when X access is configured.market-research for market sizing, competitors, or business use cases.prediction-market-risk-review before any workflow touches user capital,portfolio data, or execution-capable credentials.
Default to a compact brief containing retrieved_at, source links, source-provided timestamps, freshness caveats, facts, market-implied signals, interpretation, and actionable open questions. End with:
textThis is market intelligence, not investment or trading advice.
If access is missing, say:
textItô live basket/API data requires gated access. Request an ITO_API_KEY before using Itô-backed reads.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-13 | fail→pass | 10,943 | 17,757 | +62% | 1 | 1 | 0% | 1,814 | 3,347 | +85% | 0 | 0 | — |
case-06 | fail→pass | 10,502 | 3,915 | -63% | 1 | 1 | 0% | 2,214 | 1,620 | -27% | 0 | 0 | — |
case-12 | pass→pass | 9,090 | 7,604 | -16% | 1 | 1 | 0% | 1,646 | 2,305 | +40% | 0 | 0 | — |
case-01 | fail→fail | 10,741 | 32,447 | +202% | 1 | 1 | 0% | 2,002 | 1,389 | -31% | 0 | 0 | — |
case-02 | fail→pass | 3,423 | 8,937 | +161% | 1 | 1 | 0% | 614 | 2,175 | +254% | 0 | 0 | — |
case-03 | pass→fail | 14,119 | 23,470 | +66% | 1 | 1 | 0% | 2,737 | 4,176 | +53% | 0 | 0 | — |
case-04 | fail→pass | 5,279 | 4,729 | -10% | 1 | 1 | 0% | 886 | 1,635 | +85% | 0 | 0 | — |
case-05 | pass→pass | 9,955 | 5,618 | -44% | 1 | 1 | 0% | 1,844 | 2,113 | +15% | 0 | 0 | — |
case-07 | pass→pass | 6,991 | 4,486 | -36% | 1 | 1 | 0% | 1,071 | 1,844 | +72% | 0 | 0 | — |
case-08 | fail→fail | 6,597 | 4,755 | -28% | 1 | 1 | 0% | 233 | 1,248 | +436% | 0 | 0 | — |
case-09 | fail→fail | 11,869 | 16,488 | +39% | 1 | 1 | 0% | 2,093 | 3,263 | +56% | 0 | 0 | — |
case-10 | fail→fail | 13,984 | 4,906 | -65% | 1 | 1 | 0% | 2,855 | 1,282 | -55% | 0 | 0 | — |
case-11 | fail→fail | 8,469 | 10,213 | +21% | 1 | 1 | 0% | 1,723 | 3,043 | +77% | 0 | 0 | — |
case-14 | pass→fail | 14,926 | 24,578 | +65% | 1 | 1 | 0% | 2,509 | 4,944 | +97% | 0 | 0 | — |
case-15 | fail→pass | 7,665 | 4,508 | -41% | 1 | 1 | 0% | 1,670 | 1,924 | +15% | 0 | 0 | — |
case-16 | fail→pass | 7,119 | 1,470 | -79% | 1 | 1 | 0% | 1,320 | 1,183 | -10% | 0 | 0 | — |
case-17 | fail→fail | 10,233 | 9,383 | -8% | 1 | 1 | 0% | 1,705 | 2,611 | +53% | 0 | 0 | — |
case-18 | fail→fail | 4,645 | 5,475 | +18% | 1 | 1 | 0% | 775 | 1,905 | +146% | 0 | 0 | — |
case-19 | fail→fail | 8,927 | 16,585 | +86% | 1 | 1 | 0% | 2,058 | 3,793 | +84% | 0 | 0 | — |
case-20 | pass→pass | 15,923 | 13,647 | -14% | 1 | 1 | 0% | 3,214 | 3,604 | +12% | 0 | 0 | — |
case-21 | pass→pass | 14,290 | 12,567 | -12% | 1 | 1 | 0% | 2,552 | 3,312 | +30% | 0 | 0 | — |
case-22 | pass→pass | 11,560 | 11,169 | -3% | 1 | 1 | 0% | 2,512 | 3,529 | +40% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 19 counted toward the lift figure. The other 3 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +18 percentage points is the difference between those two pass rates over the 19 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
| Model | Method | Date | Lift |
|---|---|---|---|
| gemini-3.6-flash | verified | 7/27/2026 | +32% |
Other measured skills in the registry, with their headline benchmark lift.