Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Show this repo's context-cost leaderboard — the files that pulled the most tokens into context
.claude/skills/karanb192-leaderboard/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 26% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 7% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -71% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -36% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -43% | 0% |
!node "${CLAUDE_SKILL_DIR}/../../context-hogs.js" --render 2>/dev/null || node "${CLAUDE_PLUGIN_ROOT}/context-hogs.js" --render
The table above is this repository's most token-expensive files, aggregated across your recent sessions (read count, cumulative tokens, estimated cost). It is produced by the context-hogs hook, which records as you work.
Briefly call out the top 1-3 offenders and suggest one concrete fix each: for example: split an oversized file, exclude a generated/vendored path, or add the suggested CLAUDE.md / permissions.deny block for a file Claude keeps re-reading. Keep it short; do not re-run any commands.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-14 | fail→fail | 4,338 | 10,710 | +147% | 1 | 1 | 0% | 673 | 1,908 | +184% | 0 | 0 | — |
case-15 | fail→fail | 10,412 | 5,533 | -47% | 1 | 1 | 0% | 1,773 | 1,231 | -31% | 0 | 0 | — |
case-01 | fail→pass | 5,386 | 5,585 | +4% | 1 | 1 | 0% | 846 | 1,065 | +26% | 0 | 0 | — |
case-02 | fail→fail | 36,379 | 8,691 | -76% | 1 | 1 | 0% | 1,012 | 1,683 | +66% | 0 | 0 | — |
case-03 | fail→fail | 16,294 | 8,890 | -45% | 1 | 1 | 0% | 2,107 | 1,903 | -10% | 0 | 0 | — |
case-04 | pass→fail | 15,604 | 11,883 | -24% | 1 | 1 | 0% | 2,857 | 1,112 | -61% | 0 | 0 | — |
case-05 | pass→pass | 9,863 | 5,125 | -48% | 1 | 1 | 0% | 1,761 | 1,153 | -35% | 0 | 0 | — |
case-06 | fail→fail | 9,833 | 7,781 | -21% | 1 | 1 | 0% | 1,777 | 1,582 | -11% | 0 | 0 | — |
case-07 | fail→pass | 10,940 | 9,581 | -12% | 1 | 1 | 0% | 1,825 | 1,960 | +7% | 0 | 0 | — |
case-08 | fail→pass | 32,420 | 8,021 | -75% | 1 | 1 | 0% | 6,300 | 1,829 | -71% | 0 | 0 | — |
case-21 | pass→fail | 9,049 | 9,936 | +10% | 1 | 1 | 0% | 1,715 | 1,695 | -1% | 0 | 0 | — |
case-09 | fail→fail | 7,704 | 5,687 | -26% | 1 | 1 | 0% | 1,351 | 1,234 | -9% | 0 | 0 | — |
case-10 | fail→pass | 14,131 | 5,105 | -64% | 1 | 1 | 0% | 1,738 | 1,113 | -36% | 0 | 0 | — |
case-11 | fail→pass | 11,673 | 5,134 | -56% | 1 | 1 | 0% | 1,960 | 1,117 | -43% | 0 | 0 | — |
case-12 | fail→pass | 14,819 | 6,809 | -54% | 1 | 1 | 0% | 2,316 | 1,464 | -37% | 0 | 0 | — |
case-13 | fail→pass | 7,461 | 4,526 | -39% | 1 | 1 | 0% | 809 | 1,003 | +24% | 0 | 0 | — |
case-16 | fail→pass | 6,935 | 4,366 | -37% | 1 | 1 | 0% | 1,060 | 1,021 | -4% | 0 | 0 | — |
case-17 | fail→fail | 4,411 | 4,487 | +2% | 1 | 1 | 0% | 184 | 829 | +351% | 0 | 0 | — |
case-18 | fail→pass | 6,304 | 6,258 | -1% | 1 | 1 | 0% | 1,056 | 1,244 | +18% | 0 | 0 | — |
case-19 | fail→fail | 8,318 | 13,549 | +63% | 1 | 1 | 0% | 252 | 1,445 | +473% | 0 | 0 | — |
case-20 | pass→pass | 16,285 | 14,708 | -10% | 1 | 1 | 0% | 3,248 | 3,114 | -4% | 0 | 0 | — |
case-22 | pass→fail | 3,038 | 4,625 | +52% | 1 | 1 | 0% | 521 | 996 | +91% | 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 +27 percentage points is the difference between those two pass rates over the 19 comparable cases. 3 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 | 8/19/2026 | +18% |
Other measured skills in the registry, with their headline benchmark lift.