Evernote to IMA Migration Assistant

Problem

Evernote users often accumulate notes, clippings, and attachment metadata. Moving them into IMA can be tedious because .enex files preserve structure, body text contains noise, sensitive entries are hard to screen, and uploading many small files can be inefficient. This skill treats the migration as an offline pipeline with explicit export, conversion, cleanup, and import steps.

How It Works

The pipeline has six stages:
- Plaintext export: use evernote-backup with the official API to export a plaintext .enex file, avoiding the limitation that encrypted .notes files cannot be directly decrypted.
- Markdown conversion: enex2md.py / convert_all.py batch-convert notebooks while preserving hierarchy and generating index.md.
- Content slimming: slim_notes.py can preview with --dry-run before removing ads, boilerplate, or other noise.
- Optional merging: merge_notes.py combines many notes in one notebook into fewer files, reducing upload batches.
- Sensitive scanning: sensitive_scan.py flags possible passwords, account data, journals, or other risky content; exit code 2 requires manual review.
- IMA import: the ima-mcp connector and upload_cos.py handle create_media, COS upload, and add_knowledge.

The scripts target Python 3.6+ and mainly use the standard library.

Boundaries

This skill is suited for structured, offline Evernote backup migration, not for bypassing .notes encryption or processing unauthorized data. Keep the plaintext .enex until migration is complete, exclude sensitive notebooks from cloud import, and pass COS credentials via environment variables or stdin. Delete credential-bearing intermediate files immediately.