Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →upload a photo/image/screenshot, host an image, get a public image URL, put images on Vercel Blob, embed images in a PR/comment/doc. Upload local images to Vercel Blob and print public URLs.
.claude/skills/devin-axis-upload-photo/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -58% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -60% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 86% | 0% |
| case-22 | ✗→✓ | ▲ Improved | 272% | 0% |
| case-10 | ✓→✗ | ▼ Worse | -84% | 0% |
Upload local image files to Vercel Blob and return public URLs for sharing.
Use whenever an image or screenshot needs a public URL, including PR comments, docs, and sharing.
BLOB_READ_WRITE_TOKEN must be set in the environment. If it is missing, use the get-env-var skill to fetch it from Infisical:
bashexport BLOB_READ_WRITE_TOKEN="$(infisical secrets get BLOB_READ_WRITE_TOKEN --plain --silent)"
Run the bundled script from the repo root:
bashnode .opencode/skills/upload-photo/scripts/upload.mjs <file.png> [more files...] [--prefix <path/prefix>] [--stable]
It prints one public URL per line (https://<store>.public.blob.vercel-storage.com/...). --prefix defaults to uploads/<YYYY-MM-DD>. By default Vercel Blob appends a random suffix to the pathname for collision safety; --stable disables that (x-add-random-suffix: 0) for deterministic URLs, so overwrites are possible.
Use Markdown:
markdown
Or HTML for size control in GitHub comments:
html<img src="url" width="700">
BLOB_READ_WRITE_TOKEN.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 10,165 | 16,075 | +58% | 1 | 1 | 0% | 854 | 658 | -23% | 0 | 0 | — |
case-02 | fail→fail | 11,217 | 24,417 | +118% | 1 | 1 | 0% | 930 | 682 | -27% | 0 | 0 | — |
case-03 | fail→fail | 15,126 | 16,468 | +9% | 1 | 1 | 0% | 228 | 774 | +239% | 0 | 0 | — |
case-04 | fail→pass | 23,816 | 9,202 | -61% | 1 | 1 | 0% | 2,183 | 914 | -58% | 0 | 0 | — |
case-05 | fail→fail | 16,080 | 16,561 | +3% | 1 | 1 | 0% | 328 | 721 | +120% | 0 | 0 | — |
case-06 | fail→fail | 12,499 | 15,774 | +26% | 1 | 1 | 0% | 1,249 | 686 | -45% | 0 | 0 | — |
case-07 | fail→pass | 18,581 | 8,419 | -55% | 1 | 1 | 0% | 2,349 | 933 | -60% | 0 | 0 | — |
case-08 | fail→fail | 19,129 | 16,872 | -12% | 1 | 1 | 0% | 2,322 | 937 | -60% | 0 | 0 | — |
case-09 | fail→fail | 26,187 | 16,848 | -36% | 1 | 1 | 0% | 2,588 | 782 | -70% | 0 | 0 | — |
case-10 | pass→fail | 35,580 | 20,817 | -41% | 1 | 1 | 0% | 6,046 | 943 | -84% | 0 | 0 | — |
case-11 | fail→fail | 14,399 | 17,328 | +20% | 1 | 1 | 0% | 1,564 | 852 | -46% | 0 | 0 | — |
case-12 | fail→fail | 15,406 | 16,828 | +9% | 1 | 1 | 0% | 1,776 | 785 | -56% | 0 | 0 | — |
case-13 | pass→fail | 8,468 | 17,078 | +102% | 1 | 1 | 0% | 536 | 862 | +61% | 0 | 0 | — |
case-14 | fail→pass | 14,433 | 23,226 | +61% | 1 | 1 | 0% | 1,012 | 1,879 | +86% | 0 | 0 | — |
case-15 | fail→fail | 10,316 | 9,191 | -11% | 1 | 1 | 0% | 879 | 848 | -4% | 0 | 0 | — |
case-16 | fail→fail | 15,564 | 18,378 | +18% | 1 | 1 | 0% | 1,835 | 952 | -48% | 0 | 0 | — |
case-17 | pass→pass | 8,956 | 9,916 | +11% | 1 | 1 | 0% | 710 | 1,211 | +71% | 0 | 0 | — |
case-18 | pass→fail | 13,116 | 16,613 | +27% | 1 | 1 | 0% | 1,473 | 669 | -55% | 0 | 0 | — |
case-19 | pass→fail | 20,439 | 14,846 | -27% | 1 | 1 | 0% | 1,576 | 579 | -63% | 0 | 0 | — |
case-20 | fail→fail | 17,484 | 16,600 | -5% | 1 | 1 | 0% | 2,250 | 702 | -69% | 0 | 0 | — |
case-21 | fail→fail | 12,316 | 17,301 | +40% | 1 | 1 | 0% | 1,089 | 732 | -33% | 0 | 0 | — |
case-22 | fail→pass | 7,482 | 8,027 | +7% | 1 | 1 | 0% | 234 | 871 | +272% | 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 7 counted toward the lift figure. The other 15 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 0 percentage points is the difference between those two pass rates over the 7 comparable cases. 7 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.
Other measured skills in the registry, with their headline benchmark lift.