Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use this skill when users ask about news, media bias, stock markets, options trading, crypto, market analysis, balanced perspectives, or trending memes.
.claude/skills/bilal140202-helium-mcp-news-markets-ai-intelligence/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-17 | ✗→✓ | ▲ Improved | -43% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | -9% | 0% |
| case-05 | ✓→✓ | ＝ Same ✓ | 72% | 0% |
| case-06 | ✓→✓ | ＝ Same ✓ | -1% | 0% |
| case-01 | ✗→✗ | ＝ Same ✗ | -63% | 0% |
Use this skill when users ask about news, media bias, stock markets, options trading, crypto, market analysis, balanced perspectives, or trending memes.
Helium MCP must be configured. Add to your MCP settings:
json{ "mcpServers": { "helium": { "url": "https://heliumtrades.com/mcp" } } }
search_news with different sources, then highlight bias differencessearch_balanced_news for multi-perspective synthesis on any topicget_ticker for price + AI analysis, then get_option_price for derivativesget_top_trading_strategies for AI-ranked setups| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 17,310 | 6,691 | -61% | 1 | 1 | 0% | 2,562 | 955 | -63% | 0 | 0 | — |
case-02 | fail→fail | 20,085 | 7,610 | -62% | 1 | 1 | 0% | 3,085 | 844 | -73% | 0 | 0 | — |
case-03 | fail→fail | 19,785 | 6,413 | -68% | 1 | 1 | 0% | 3,543 | 899 | -75% | 0 | 0 | — |
case-04 | pass→pass | 13,159 | 10,326 | -22% | 1 | 1 | 0% | 2,828 | 2,580 | -9% | 0 | 0 | — |
case-05 | pass→pass | 2,593 | 2,333 | -10% | 1 | 1 | 0% | 512 | 880 | +72% | 0 | 0 | — |
case-06 | pass→pass | 9,676 | 5,917 | -39% | 1 | 1 | 0% | 1,738 | 1,718 | -1% | 0 | 0 | — |
case-07 | fail→fail | 6,529 | 2,212 | -66% | 1 | 1 | 0% | 1,110 | 790 | -29% | 0 | 0 | — |
case-08 | fail→fail | 8,948 | 5,217 | -42% | 1 | 1 | 0% | 1,359 | 1,249 | -8% | 0 | 0 | — |
case-09 | fail→fail | 13,114 | 4,615 | -65% | 1 | 1 | 0% | 2,108 | 637 | -70% | 0 | 0 | — |
case-10 | fail→fail | 16,386 | 6,684 | -59% | 1 | 1 | 0% | 2,860 | 645 | -77% | 0 | 0 | — |
case-11 | fail→fail | 16,639 | 2,057 | -88% | 1 | 1 | 0% | 2,518 | 778 | -69% | 0 | 0 | — |
case-12 | fail→fail | 6,408 | 5,229 | -18% | 1 | 1 | 0% | 980 | 755 | -23% | 0 | 0 | — |
case-13 | fail→fail | 14,001 | 2,841 | -80% | 1 | 1 | 0% | 2,755 | 928 | -66% | 0 | 0 | — |
case-14 | fail→fail | 9,544 | 4,684 | -51% | 1 | 1 | 0% | 1,670 | 683 | -59% | 0 | 0 | — |
case-15 | fail→fail | 14,492 | 17,979 | +24% | 1 | 1 | 0% | 2,301 | 694 | -70% | 0 | 0 | — |
case-16 | fail→fail | 13,014 | 2,107 | -84% | 1 | 1 | 0% | 2,363 | 796 | -66% | 0 | 0 | — |
case-17 | fail→pass | 10,190 | 3,601 | -65% | 1 | 1 | 0% | 1,756 | 1,000 | -43% | 0 | 0 | — |
case-18 | fail→fail | 15,198 | 4,159 | -73% | 1 | 1 | 0% | 2,483 | 657 | -74% | 0 | 0 | — |
case-19 | fail→fail | 13,137 | 4,545 | -65% | 1 | 1 | 0% | 2,288 | 675 | -70% | 0 | 0 | — |
case-20 | fail→fail | 17,947 | 2,475 | -86% | 1 | 1 | 0% | 2,779 | 819 | -71% | 0 | 0 | — |
case-21 | fail→fail | 14,540 | 4,500 | -69% | 1 | 1 | 0% | 2,535 | 640 | -75% | 0 | 0 | — |
case-22 | fail→fail | 14,233 | 3,348 | -76% | 1 | 1 | 0% | 2,471 | 810 | -67% | 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 11 counted toward the lift figure. The other 11 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 +5 percentage points is the difference between those two pass rates over the 11 comparable cases.
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.
Other measured skills in the registry, with their headline benchmark lift.