reMarkable Tablet Sync

Bidirectional sync with reMarkable tablet via Cloud API (rmapi). Fetch handwritten notes/sketches, process with AI, and push content back. Use for sketch enhancement, journal extraction, or sending documents/images to the tablet.

Security Scan

The skill's requirements and instructions are consistent with its stated purpose (reMarkable sync via rmapi); no unrelated credentials or surprising installs are requested — main residual risks are trusting the downloaded rmapi binary and that content is synchronized via the reMarkable cloud/token stored in ~/.rmapi.

Purpose & Capability

Name/description (bidirectional rmapi sync, fetch/process/push notes) matches the instructions: it guides installing rmapi, converting .rm to SVG/PNG, running OCR/AI image edits, and pushing PDFs back. No unrelated services, credentials, or binaries are requested.

Instruction Scope

SKILL.md stays on-task: lists rmapi commands (ls, get, put, mget/mput), unzip/find/rmc/cairosvg/Pillow workflows, and an authentication step that uses my.remarkable.com connect flow. It does not instruct reading arbitrary system files or unrelated environment variables. It does note rmapi saves a token to ~/.rmapi (expected).

Install Mechanism

No registry-level install spec (instruction-only). The README recommends downloading a rmapi binary from a GitHub release and pip installing Python packages (rmc, cairosvg, pillow). GitHub releases are a common source, but downloading and executing a binary from a third-party fork (ddvk) requires user trust; verify checksums/releases and prefer official upstream if desired.

Credentials

The skill declares no required env vars or secrets. The only persisted credential is the rmapi token saved by rmapi to ~/.rmapi after the interactive login, which is the expected mechanism for this tool and proportional to the task.

Persistence & Privilege

always is false and the skill is user-invocable. Model invocation is allowed (platform default) but that is not combined with any broad or hidden privileges. The skill does not request permanent system-wide changes beyond using rmapi's own token storage.

Assessment

This skill is internally consistent with its purpose, but before installing: 1) verify/trust the rmapi binary you download (prefer official releases or verify checksums) because you will execute it; 2) be aware that rmapi uses the reMarkable cloud and saves an auth token to ~/.rmapi — any content synced will pass through reMarkable’s cloud; 3) pip installs are run under your user account (use virtualenvs if you prefer); 4) the instructions expect you to run conversion commands and Python snippets locally—review them before executing; and 5) if you allow autonomous agent actions, remember the agent could run the described commands (download, convert, upload) — only grant that capability if you trust the skill and agent.

reMarkable Tablet Integration (rmapi)

Bidirectional sync with reMarkable tablet via Cloud API. Fetch sketches and notes, process them (AI enhancement, text extraction), and push content back to the tablet.

Typical Use Cases

1. Sketch → AI → Tablet loop — Fetch rough sketch → enhance with AI → push polished version back
2. Journal entries — Fetch handwritten thoughts → interpret → append to memory/journal
3. Brainstorming — Fetch diagrams/lists → extract structure → add to project docs
4. Send reading material — Push PDFs/documents to tablet for offline reading
5. AI art to tablet — Generate images → convert to PDF → push for viewing on e-ink

Bidirectional Pipeline

┌─────────────────────────────────────────────────────────────────┐
│                        FETCH (tablet → agent)                    │
├─────────────────────────────────────────────────────────────────┤
│  reMarkable → Cloud sync → rmapi get → .rmdoc                   │
│                                           ↓                      │
│                              unzip → .rm file → rmc → SVG       │
│                                                      ↓           │
│                                          cairosvg → PNG          │
│                                                      ↓           │
│                               ┌──────────┴──────────┐            │
│                          Text content?         Visual/sketch?    │
│                               ↓                      ↓           │
│                        OCR/interpret          AI image editing   │
│                               ↓                      ↓           │
│                        Add to memory        Enhanced image       │
└─────────────────────────────────────────────────────────────────┘

┌─────────────────────────────────────────────────────────────────┐
│                        PUSH (agent → tablet)                     │
├─────────────────────────────────────────────────────────────────┤
│  Image/document → Convert to PDF (if needed) → rmapi put        │
│                                                      ↓           │
│                                           Cloud sync → tablet    │
└─────────────────────────────────────────────────────────────────┘

Prerequisites

1. Install rmapi

Download the latest release from juruen/rmapi (ddvk fork recommended):

# Example for Linux amd64
curl -L https://github.com/ddvk/rmapi/releases/latest/download/rmapi-linux-amd64 -o ~/bin/rmapi
chmod +x ~/bin/rmapi

2. Install conversion tools

pip install --user rmc cairosvg pillow

• rmc — Converts .rm stroke files to SVG
• cairosvg — Converts SVG to PNG
• pillow — Converts PNG to PDF for pushing back

3. Authenticate (ONE-TIME)

1. Go to https://my.remarkable.com/connect/desktop
2. Log in and copy the 8-character code
3. Run rmapi and paste the code when prompted
4. Token saved to ~/.rmapi — future runs are automatic

Commands

Fetch (Download from tablet)

# List files/folders
rmapi ls
rmapi ls -l -t              # Long format, sorted by time

# Refresh from cloud
rmapi refresh

# Navigate
rmapi cd "folder name"

# Find files
rmapi find --starred /              # Starred items
rmapi find --tag="my-tag" /         # By tag
rmapi find / ".*sketch.*"           # By regex

# Download single file
rmapi get "filename"
rmapi get "Folder/Notebook"

# Download with annotations (best for handwritten content)
rmapi geta "filename"

# Bulk download folder
rmapi mget -o ./sync-folder/ "/My Folder"

Push (Upload to tablet)

# Upload single file (PDF or EPUB only)
rmapi put document.pdf
rmapi put document.pdf "Target Folder/"

# Bulk upload
rmapi mput ./local-folder/ "Remote Folder/"

# Create folder on tablet
rmapi mkdir "New Folder"

Supported formats: PDF, EPUB

Conversion Workflows

Fetch: .rmdoc → PNG

reMarkable notebooks download as .rmdoc (a ZIP archive containing stroke data):

# 1. Download the notebook
rmapi get "Folder/MyNotebook"

# 2. Extract (it's a zip)
unzip "MyNotebook.rmdoc" -d extracted/

# 3. Find the .rm stroke file(s)
# Structure: extracted/<doc-uuid>/<page-uuid>.rm
find extracted -name "*.rm"

# 4. Convert .rm → SVG
rmc -t svg -o page.svg "extracted/<doc-uuid>/<page-uuid>.rm"

# 5. Convert SVG → PNG (reMarkable dimensions: 1404×1872)
python3 -c "
import cairosvg
cairosvg.svg2png(url='page.svg', write_to='page.png', output_width=1404, output_height=1872)
"

Push: Image → PDF

reMarkable only accepts PDF/EPUB, so convert images first:

from PIL import Image

img = Image.open('image.png')
rgb = img.convert('RGB')
rgb.save('image.pdf', 'PDF', resolution=150)

Then push:

rmapi put image.pdf "My Folder/"

Full Workflow Example

Sketch enhancement loop:

# 1. Fetch sketch from tablet
rmapi get "Sketches/MyDrawing"

# 2. Extract and convert to PNG
unzip MyDrawing.rmdoc -d MyDrawing_extracted/
RM_FILE=$(find MyDrawing_extracted -name "*.rm" | head -1)
rmc -t svg -o sketch.svg "$RM_FILE"
python3 -c "import cairosvg; cairosvg.svg2png(url='sketch.svg', write_to='sketch.png', output_width=1404, output_height=1872)"

# 3. [Your AI enhancement step here]
# Example: use any image-to-image AI tool to enhance sketch.png → enhanced.png

# 4. Convert to PDF and push back
python3 -c "from PIL import Image; Image.open('enhanced.png').convert('RGB').save('enhanced.pdf', 'PDF', resolution=150)"
rmapi put enhanced.pdf "Sketches/"

Sharing Strategies

Create a dedicated sync folder on your tablet, or use:

• Tags — Tag items for discovery with rmapi find --tag
• Stars — Star items for quick access with rmapi find --starred

Notes

• Cloud sync required — Tablet must sync to cloud before files are available (pull down to refresh on tablet)
• Format — .rmdoc is a ZIP containing JSON metadata + .rm binary stroke files
• Warnings — rmc may show warnings about newer format versions — usually still works
• Dimensions — reMarkable display is 1404×1872 pixels (portrait)
• Text extraction — For handwritten text, use vision models to interpret visually rather than traditional OCR

Troubleshooting

Issue	Solution
rmapi not connecting	Re-authenticate: delete ~/.rmapi and run rmapi again
File not found after upload	Wait for cloud sync, or refresh tablet manually
rmc format warnings	Usually safe to ignore; output still generated
SVG looks empty	Check if the correct .rm file was used (multi-page notebooks have multiple)