Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Optimize Evernote integration costs and resource usage. Use when managing API quotas, reducing storage usage, or optimizing upload limits. Trigger with phrases like "evernote cost", "evernote quota", "evernote limits", "evernote upload".
.claude/skills/jeremylongshore-evernote-cost-tuning/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 34% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 50% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 27% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 73% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 4% | 0% |
Optimize resource usage and manage costs in Evernote integrations, focusing on monthly upload quotas, storage efficiency, image compression, and account limit monitoring.
user.accountingQuery userStore.getUser() to access user.accounting which contains uploadLimit, uploaded, and uploadLimitEnd. Calculate remaining quota and percentage used.
javascriptasync function getQuotaStatus(userStore) { const user = await userStore.getUser(); const { uploadLimit, uploaded, uploadLimitEnd } = user.accounting; return { totalMB: Math.round(uploadLimit / 1024 / 1024), usedMB: Math.round(uploaded / 1024 / 1024), remainingMB: Math.round((uploadLimit - uploaded) / 1024 / 1024), percentUsed: Math.round((uploaded / uploadLimit) * 100), resetsAt: new Date(uploadLimitEnd) }; }
Compress images before attaching to notes. Resize large images to a maximum dimension (e.g., 1920px). Convert PNG screenshots to JPEG for smaller file sizes. Skip attaching files that exceed the single-note size limit (25MB for Basic, 200MB for Premium).
javascriptfunction estimateNoteSize(content, resources = []) { const contentBytes = Buffer.byteLength(content, 'utf8'); const resourceBytes = resources.reduce((sum, r) => sum + r.data.size, 0); return contentBytes + resourceBytes; } function canUpload(noteSize, remainingQuota) { return noteSize < remainingQuota; }
Check quota before creating notes with large attachments. Use findNotesMetadata() for read operations (zero upload cost). Batch small notes into single notes where appropriate.
Find large notes consuming quota. List notes sorted by content length to identify optimization candidates. Remove unused resources and delete notes in trash to reclaim space.
Send alerts when upload usage exceeds thresholds (e.g., 75%, 90%, 95%). Log quota status after each upload operation for trend analysis.
For the complete quota monitor, image optimizer, cleanup utilities, and alert system, see Implementation Guide.
| Limit | Basic | Premium | Business | |-------|-------|---------|----------| | Monthly upload | 60 MB | 10 GB | 20 GB/user | | Single note size | 25 MB | 200 MB | 200 MB | | Notebooks | 250 | 250 | 10,000 | | Tags | 100,000 | 100,000 | 100,000 | | Notes | 100,000 | 100,000 | 500,000 |
| Error | Cause | Solution | |-------|-------|----------| | QUOTA_REACHED | Monthly upload limit exceeded | Wait for quota reset at uploadLimitEnd | | LIMIT_REACHED | Too many notebooks or tags | Delete unused notebooks, merge duplicate tags | | DATA_REQUIRED | Empty note after optimization | Ensure note still has content after compression | | BAD_DATA_FORMAT | Attachment hash mismatch | Recompute MD5 hash after image compression |
For architecture patterns, see evernote-reference-architecture.
Quota dashboard: Build a dashboard showing current upload usage (MB used / total), days until reset, largest notes by size, and projected usage based on recent trends.
Image pipeline: Before attaching images, resize to max 1920px width, convert PNG to JPEG at 80% quality, check resulting size against remaining quota, and skip if insufficient.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 23,373 | 15,614 | -33% | 1 | 1 | 0% | 3,793 | 4,145 | +9% | 0 | 0 | — |
case-02 | fail→fail | 24,530 | 18,781 | -23% | 1 | 1 | 0% | 4,004 | 4,149 | +4% | 0 | 0 | — |
case-03 | fail→pass | 17,981 | 15,159 | -16% | 1 | 1 | 0% | 2,545 | 3,422 | +34% | 0 | 0 | — |
case-04 | fail→pass | 15,183 | 13,873 | -9% | 1 | 1 | 0% | 2,548 | 3,817 | +50% | 0 | 0 | — |
case-05 | fail→fail | 16,268 | 8,700 | -47% | 1 | 1 | 0% | 2,236 | 2,634 | +18% | 0 | 0 | — |
case-06 | pass→pass | 11,168 | 12,204 | +9% | 1 | 1 | 0% | 1,854 | 3,242 | +75% | 0 | 0 | — |
case-07 | pass→pass | 6,901 | 5,383 | -22% | 1 | 1 | 0% | 1,029 | 1,859 | +81% | 0 | 0 | — |
case-08 | pass→pass | 9,801 | 9,424 | -4% | 1 | 1 | 0% | 1,612 | 2,175 | +35% | 0 | 0 | — |
case-09 | fail→pass | 14,712 | 9,723 | -34% | 1 | 1 | 0% | 2,018 | 2,564 | +27% | 0 | 0 | — |
case-10 | pass→pass | 9,613 | 5,482 | -43% | 1 | 1 | 0% | 1,409 | 2,096 | +49% | 0 | 0 | — |
case-11 | fail→pass | 24,592 | 6,250 | -75% | 1 | 1 | 0% | 1,272 | 2,203 | +73% | 0 | 0 | — |
case-12 | pass→pass | 6,862 | 2,012 | -71% | 1 | 1 | 0% | 850 | 1,416 | +67% | 0 | 0 | — |
case-13 | pass→pass | 6,638 | 2,080 | -69% | 1 | 1 | 0% | 1,061 | 1,442 | +36% | 0 | 0 | — |
case-14 | fail→pass | 7,866 | 1,997 | -75% | 1 | 1 | 0% | 1,348 | 1,404 | +4% | 0 | 0 | — |
case-15 | fail→pass | 5,929 | 5,432 | -8% | 1 | 1 | 0% | 962 | 1,785 | +86% | 0 | 0 | — |
case-16 | pass→pass | 5,639 | 1,398 | -75% | 1 | 1 | 0% | 841 | 1,296 | +54% | 0 | 0 | — |
case-17 | fail→pass | 13,658 | 5,223 | -62% | 1 | 1 | 0% | 1,930 | 1,678 | -13% | 0 | 0 | — |
case-18 | pass→pass | 16,596 | 14,383 | -13% | 1 | 1 | 0% | 3,198 | 3,762 | +18% | 0 | 0 | — |
case-19 | pass→pass | 15,418 | 18,439 | +20% | 1 | 1 | 0% | 2,484 | 4,420 | +78% | 0 | 0 | — |
case-20 | pass→pass | 6,881 | 7,765 | +13% | 1 | 1 | 0% | 863 | 2,180 | +153% | 0 | 0 | — |
case-21 | pass→pass | 18,732 | 14,905 | -20% | 1 | 1 | 0% | 3,823 | 3,900 | +2% | 0 | 0 | — |
case-22 | pass→pass | 12,487 | 9,462 | -24% | 1 | 1 | 0% | 1,850 | 2,482 | +34% | 0 | 0 | — |
case-23 | pass→pass | 18,253 | 16,541 | -9% | 1 | 1 | 0% | 2,632 | 3,561 | +35% | 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. 23 cases were attempted, and 22 counted toward the lift figure. The other 1 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 +30 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.