Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Searches GitHub for existing implementations, libraries, or patterns. Use when finding code examples or prior art on a topic during a research session.
.claude/skills/athola-code-search/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-15 | ✗→✓ | ▲ Improved | 106% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 255% | 0% |
| case-17 | ✗→✓ | ▲ Improved | -53% | 0% |
| case-21 | ✗→✓ | ▲ Improved | -52% | 0% |
| case-07 | ✓→✗ | ▼ Worse | -55% | 0% |
/tome:research session or standalone searchSkill(tome:papers))Search GitHub for implementations of a given topic.
Invoked as part of /tome:research or standalone.
tome.channels.github.build_github_search_queries()
parse_github_result()build_github_api_search() for richer metadata
rank_github_findings()WebSearch for the requested topic
(name, URL, stars) and a relevance score
rank_github_findings() sorted byrelevance score descending
explicitly rather than returning an empty list silently
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 48,932 | 46,545 | -5% | 1 | 1 | 0% | 3,644 | 3,819 | +5% | 0 | 0 | — |
case-02 | fail→fail | 17,195 | 5,887 | -66% | 1 | 1 | 0% | 3,511 | 814 | -77% | 0 | 0 | — |
case-03 | fail→fail | 12,385 | 22,150 | +79% | 1 | 1 | 0% | 2,586 | 4,933 | +91% | 0 | 0 | — |
case-04 | fail→fail | 5,152 | 2,251 | -56% | 1 | 1 | 0% | 356 | 600 | +69% | 0 | 0 | — |
case-05 | fail→fail | 22,041 | 52,307 | +137% | 1 | 1 | 0% | 3,960 | 4,309 | +9% | 0 | 0 | — |
case-06 | fail→fail | 3,735 | 7,486 | +100% | 1 | 1 | 0% | 455 | 662 | +45% | 0 | 0 | — |
case-07 | pass→fail | 7,718 | 5,977 | -23% | 1 | 1 | 0% | 1,411 | 637 | -55% | 0 | 0 | — |
case-08 | fail→fail | 21,991 | 4,972 | -77% | 1 | 1 | 0% | 3,179 | 617 | -81% | 0 | 0 | — |
case-09 | fail→fail | 3,080 | 11,986 | +289% | 1 | 1 | 0% | 355 | 766 | +116% | 0 | 0 | — |
case-10 | fail→fail | 3,551 | 13,431 | +278% | 1 | 1 | 0% | 562 | 2,518 | +348% | 0 | 0 | — |
case-11 | fail→fail | 26,280 | 8,960 | -66% | 1 | 1 | 0% | 3,818 | 958 | -75% | 0 | 0 | — |
case-12 | pass→fail | 14,346 | 6,861 | -52% | 1 | 1 | 0% | 2,407 | 547 | -77% | 0 | 0 | — |
case-13 | fail→fail | 8,211 | 6,069 | -26% | 1 | 1 | 0% | 1,388 | 822 | -41% | 0 | 0 | — |
case-14 | pass→pass | 12,839 | 17,538 | +37% | 1 | 1 | 0% | 2,024 | 4,024 | +99% | 0 | 0 | — |
case-15 | fail→pass | 13,585 | 26,847 | +98% | 1 | 1 | 0% | 2,471 | 5,091 | +106% | 0 | 0 | — |
case-16 | fail→pass | 10,082 | 37,848 | +275% | 1 | 1 | 0% | 1,883 | 6,687 | +255% | 0 | 0 | — |
case-17 | fail→pass | 10,932 | 4,248 | -61% | 1 | 1 | 0% | 2,026 | 954 | -53% | 0 | 0 | — |
case-18 | fail→fail | 16,686 | 5,991 | -64% | 1 | 1 | 0% | 3,300 | 559 | -83% | 0 | 0 | — |
case-19 | pass→fail | 17,046 | 5,432 | -68% | 1 | 1 | 0% | 2,952 | 618 | -79% | 0 | 0 | — |
case-20 | fail→fail | 9,646 | 9,156 | -5% | 1 | 1 | 0% | 1,777 | 1,502 | -15% | 0 | 0 | — |
case-21 | fail→pass | 18,587 | 5,981 | -68% | 1 | 1 | 0% | 2,660 | 1,281 | -52% | 0 | 0 | — |
case-22 | fail→fail | 6,926 | 7,062 | +2% | 1 | 1 | 0% | 343 | 699 | +104% | 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. 4 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 | 8/21/2026 | -5% |
Other measured skills in the registry, with their headline benchmark lift.