Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Improves the quality of images, especially screenshots, by enhancing resolution, sharpness, and clarity. Perfect for preparing images for presentations, documentation, or social media posts.
.claude/skills/composiohq-image-enhancer/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 2% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -7% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -4% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -36% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 31% | 0% |
This skill takes your images and screenshots and makes them look better—sharper, clearer, and more professional.
Improve the image quality of screenshot.pngEnhance all images in this folderUpscale this image to 4K resolutionSharpen this blurry screenshotReduce compression artifacts in this imageImprove the quality of all PNG files in this directoryUser: "Improve the image quality of screenshot-2024.png"
Output:
Analyzing screenshot-2024.png...
Current specs:
- Resolution: 1920x1080
- Format: PNG
- Quality: Good, but slight blur
Enhancements applied:
✓ Upscaled to 2560x1440 (retina)
✓ Sharpened edges
✓ Enhanced text clarity
✓ Optimized file size
Saved as: screenshot-2024-enhanced.png
Original preserved as: screenshot-2024-original.pngInspired by: Lenny Rachitsky's workflow from his newsletter - used for screenshots in his articles
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-05 | fail→fail | 24,628 | 9,000 | -63% | 1 | 1 | 0% | 1,944 | 2,012 | +3% | 0 | 0 | — |
case-06 | pass→pass | 11,344 | 7,371 | -35% | 1 | 1 | 0% | 1,950 | 1,628 | -17% | 0 | 0 | — |
case-01 | fail→pass | 7,451 | 4,653 | -38% | 1 | 1 | 0% | 1,322 | 1,342 | +2% | 0 | 0 | — |
case-02 | fail→pass | 10,560 | 7,174 | -32% | 1 | 1 | 0% | 1,945 | 1,805 | -7% | 0 | 0 | — |
case-03 | fail→pass | 10,474 | 6,228 | -41% | 1 | 1 | 0% | 1,696 | 1,628 | -4% | 0 | 0 | — |
case-04 | pass→pass | 10,663 | 9,545 | -10% | 1 | 1 | 0% | 1,637 | 1,879 | +15% | 0 | 0 | — |
case-07 | pass→pass | 12,050 | 8,432 | -30% | 1 | 1 | 0% | 2,097 | 1,989 | -5% | 0 | 0 | — |
case-08 | fail→pass | 16,677 | 6,122 | -63% | 1 | 1 | 0% | 2,557 | 1,642 | -36% | 0 | 0 | — |
case-09 | fail→pass | 6,312 | 4,803 | -24% | 1 | 1 | 0% | 992 | 1,296 | +31% | 0 | 0 | — |
case-10 | fail→pass | 12,871 | 4,921 | -62% | 1 | 1 | 0% | 1,125 | 1,287 | +14% | 0 | 0 | — |
case-11 | pass→pass | 7,437 | 6,890 | -7% | 1 | 1 | 0% | 1,525 | 1,664 | +9% | 0 | 0 | — |
case-12 | fail→pass | 11,431 | 9,264 | -19% | 1 | 1 | 0% | 2,009 | 2,171 | +8% | 0 | 0 | — |
case-13 | fail→pass | 12,655 | 6,044 | -52% | 1 | 1 | 0% | 2,243 | 1,561 | -30% | 0 | 0 | — |
case-14 | pass→pass | 12,387 | 4,216 | -66% | 1 | 1 | 0% | 2,133 | 1,312 | -38% | 0 | 0 | — |
case-15 | fail→fail | 13,036 | 8,592 | -34% | 1 | 1 | 0% | 2,072 | 1,916 | -8% | 0 | 0 | — |
case-16 | fail→pass | 14,562 | 3,501 | -76% | 1 | 1 | 0% | 1,598 | 1,254 | -22% | 0 | 0 | — |
case-17 | fail→fail | 6,259 | 5,563 | -11% | 1 | 1 | 0% | 1,310 | 1,277 | -3% | 0 | 0 | — |
case-18 | pass→pass | 17,542 | 6,346 | -64% | 1 | 1 | 0% | 2,960 | 1,565 | -47% | 0 | 0 | — |
case-19 | pass→pass | 10,394 | 2,745 | -74% | 1 | 1 | 0% | 1,767 | 981 | -44% | 0 | 0 | — |
case-20 | fail→fail | 1,759 | 3,708 | +111% | 1 | 1 | 0% | 225 | 1,127 | +401% | 0 | 0 | — |
case-21 | fail→pass | 27,337 | 36,295 | +33% | 1 | 1 | 0% | 6,173 | 6,720 | +9% | 0 | 0 | — |
case-22 | pass→pass | 7,227 | 7,011 | -3% | 1 | 1 | 0% | 1,274 | 1,641 | +29% | 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. The headline lift of +45 percentage points is the difference between those two pass rates over the 22 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.