Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Advanced Evernote SDK patterns and best practices. Use when implementing complex note operations, batch processing, search queries, or optimizing SDK usage. Trigger with phrases like "evernote sdk patterns", "evernote best practices", "evernote advanced", "evernote batch operations".
.claude/skills/jeremylongshore-evernote-sdk-patterns/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-12 | ✗→✓ | ▲ Improved | 15% | 0% |
| case-18 | ✗→✓ | ▲ Improved | 25% | 0% |
| case-19 | ✗→✓ | ▲ Improved | 111% | 0% |
| case-01 | ✓→✓ | = Same ✓ | 32% | 0% |
| case-02 | ✓→✓ | = Same ✓ | 65% | 0% |
Production-ready patterns for working with the Evernote SDK, including search with NoteFilter, pagination, attachments, tags, error handling wrappers, and batch operations with rate limit handling.
evernote-install-auth and evernote-hello-worldUse NoteFilter for query terms and sort order, paired with NotesMetadataResultSpec to select returned fields. This avoids fetching full note content when only metadata is needed.
javascriptconst filter = new Evernote.NoteStore.NoteFilter({ words: 'tag:important notebook:Work', ascending: false, order: Evernote.Types.NoteSortOrder.UPDATED }); const spec = new Evernote.NoteStore.NotesMetadataResultSpec({ includeTitle: true, includeUpdated: true, includeTagGuids: true, includeNotebookGuid: true }); const result = await noteStore.findNotesMetadata(filter, 0, 100, spec);
Compute the MD5 hash of the file buffer, create a Resource with the binary data and MIME type, embed it in ENML with <en-media type="..." hash="..."/>, and attach it to the note.
javascriptconst hash = crypto.createHash('md5').update(fileBuffer).digest('hex'); const resource = new Evernote.Types.Resource(); resource.data = new Evernote.Types.Data(); resource.data.body = fileBuffer; resource.mime = 'image/png'; const note = new Evernote.Types.Note(); note.title = 'Note with Attachment'; note.content = wrapInENML(`<en-media type="image/png" hash="${hash}"/>`); note.resources = [resource]; await noteStore.createNote(note);
Wrap API calls to distinguish EDAMUserException (client errors), EDAMSystemException (rate limits, maintenance), and EDAMNotFoundException (invalid GUIDs). Use error.rateLimitDuration for automatic retry delays.
Process items sequentially with configurable delay between operations. On rate limit errors, wait for rateLimitDuration seconds then retry. Track progress with callbacks.
Implement getOrCreateTag() and getOrCreateNotebook() for idempotent operations. Use listTags() / listNotebooks() to check existence before creating.
For all nine patterns with complete implementations, see Implementation Guide.
NoteFilter and NotesMetadataResultSpecEvernoteError wrapper class with isRateLimit, isNotFound, isInvalidData| Error | Cause | Solution | |-------|-------|----------| | RATE_LIMIT_REACHED | Too many API calls | Use rateLimitDuration, add delays between batch items | | BAD_DATA_FORMAT | Invalid ENML | Validate with wrapInENML() before sending | | DATA_CONFLICT | Concurrent modification | Refetch note metadata and retry update | | QUOTA_REACHED | Account storage full | Check remaining quota via user.accounting |
See evernote-core-workflow-a for note creation and management workflows.
Bulk tagging: Search for all notes matching a query, then batch-add a tag to each result with 200ms delay between operations and automatic rate limit retry.
Attachment upload: Read a PDF from disk, compute its MD5 hash, create a note with the PDF as an <en-media> resource, and verify the upload via getNote() with withResources: true.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 13,730 | 11,507 | -16% | 1 | 1 | 0% | 2,121 | 2,793 | +32% | 0 | 0 | — |
case-02 | pass→pass | 5,371 | 3,525 | -34% | 1 | 1 | 0% | 969 | 1,595 | +65% | 0 | 0 | — |
case-03 | pass→pass | 12,443 | 8,935 | -28% | 1 | 1 | 0% | 1,905 | 2,446 | +28% | 0 | 0 | — |
case-04 | pass→pass | 8,423 | 6,536 | -22% | 1 | 1 | 0% | 1,208 | 1,983 | +64% | 0 | 0 | — |
case-05 | pass→pass | 10,999 | 8,973 | -18% | 1 | 1 | 0% | 1,694 | 2,789 | +65% | 0 | 0 | — |
case-06 | pass→pass | 9,776 | 3,550 | -64% | 1 | 1 | 0% | 1,354 | 1,676 | +24% | 0 | 0 | — |
case-07 | fail→fail | 11,207 | 5,949 | -47% | 1 | 1 | 0% | 1,645 | 2,152 | +31% | 0 | 0 | — |
case-08 | pass→pass | 9,245 | 9,064 | -2% | 1 | 1 | 0% | 1,774 | 2,423 | +37% | 0 | 0 | — |
case-09 | pass→pass | 12,257 | 19,497 | +59% | 1 | 1 | 0% | 2,185 | 2,992 | +37% | 0 | 0 | — |
case-10 | pass→pass | 18,992 | 7,606 | -60% | 1 | 1 | 0% | 2,781 | 2,478 | -11% | 0 | 0 | — |
case-11 | pass→pass | 18,056 | 12,107 | -33% | 1 | 1 | 0% | 2,778 | 2,939 | +6% | 0 | 0 | — |
case-12 | fail→pass | 14,509 | 12,369 | -15% | 1 | 1 | 0% | 2,617 | 3,016 | +15% | 0 | 0 | — |
case-13 | pass→pass | 11,241 | 8,998 | -20% | 1 | 1 | 0% | 2,018 | 2,650 | +31% | 0 | 0 | — |
case-14 | pass→pass | 12,573 | 10,005 | -20% | 1 | 1 | 0% | 2,035 | 2,600 | +28% | 0 | 0 | — |
case-15 | pass→pass | 8,274 | 4,748 | -43% | 1 | 1 | 0% | 1,492 | 1,790 | +20% | 0 | 0 | — |
case-16 | pass→pass | 4,246 | 2,922 | -31% | 1 | 1 | 0% | 608 | 1,399 | +130% | 0 | 0 | — |
case-17 | pass→pass | 11,046 | 3,328 | -70% | 1 | 1 | 0% | 1,862 | 1,568 | -16% | 0 | 0 | — |
case-18 | fail→pass | 12,709 | 9,516 | -25% | 1 | 1 | 0% | 2,226 | 2,779 | +25% | 0 | 0 | — |
case-19 | fail→pass | 4,200 | 3,126 | -26% | 1 | 1 | 0% | 756 | 1,593 | +111% | 0 | 0 | — |
case-20 | pass→pass | 4,976 | 3,092 | -38% | 1 | 1 | 0% | 636 | 1,547 | +143% | 0 | 0 | — |
case-21 | pass→pass | 10,791 | 9,349 | -13% | 1 | 1 | 0% | 1,837 | 2,323 | +26% | 0 | 0 | — |
case-22 | pass→pass | 19,151 | 13,614 | -29% | 1 | 1 | 0% | 3,071 | 3,658 | +19% | 0 | 0 | — |
case-23 | pass→pass | 26,559 | 12,851 | -52% | 1 | 1 | 0% | 4,937 | 3,369 | -32% | 0 | 0 | — |
case-24 | pass→pass | 20,634 | 27,075 | +31% | 1 | 1 | 0% | 3,516 | 5,258 | +50% | 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. 24 cases were attempted. The headline lift of +13 percentage points is the difference between those two pass rates over the 24 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.