Remarkable

---

name: remarkable description: Fetch handwritten notes, sketches, and drawings from a reMarkable tablet via Cloud API (rmapi). Process content by refining artwork with AI image generation, extracting handwritten text to memory/journal, or using sketches as input for other workflows. Use when working with reMarkable tablet content, syncing handwritten notes, processing sketches, or integrating tablet drawings into projects.

---

reMarkable Tablet Integration (rmapi)

Fetch handwritten notes, sketches, and drawings from a reMarkable tablet via Cloud API. Process them — refine artwork with AI image generation, extract text to memory/journal, or use as input for other workflows.

Typical Use Cases

1. Journal entries — User writes thoughts on reMarkable → fetch → OCR/interpret → append to memory/YYYY-MM-DD.md or a dedicated journal file 2. Sketch refinement — User draws a rough graphic → fetch → enhance with nano-banana-pro (AI image editing) → return polished version 3. Brainstorming/notes — User jots down ideas, lists, diagrams → fetch → extract structure → add to project docs or memory 4. Illustrations — User creates hand-drawn art → fetch → optionally stylize → use in blog posts, social media, etc.

Processing Pipeline

reMarkable tablet → Cloud sync → rmapi fetch → PDF/PNG
                                                  ↓
                                    ┌─────────────┴─────────────┐
                                    │                           │
                              Text content?               Visual/sketch?
                                    │                           │
                              OCR / interpret            nano-banana-pro
                                    │                     (AI enhance)
                                    │                           │
                              Add to memory/            Return refined
                              journal/project            image to user

Setup

Tool: rmapi (ddvk fork) v0.0.32

Binary: ~/bin/rmapi

Config: ~/.rmapi (device token after auth)

Sync folder: ~/clawd/remarkable-sync/

Authentication (ONE-TIME)

1. User goes to https://my.remarkable.com/connect/desktop 2. Logs in, gets 8-character code 3. Run rmapi and enter the code 4. Token saved to ~/.rmapi — future runs are automatic

Commands

# List files/folders
rmapi ls
rmapi ls --json
Navigate
rmapi cd "folder name"
Find by tag / starred / regex
rmapi find --tag="share-with-gandalf" /
rmapi find --starred /
rmapi find / ".*sketch.*"
Download (PDF)
rmapi get "filename"
Download with annotations rendered (best for sketches)
rmapi geta "filename"
Bulk download folder
rmapi mget -o ~/clawd/remarkable-sync/ "/Shared with Gandalf"

Sharing Workflows

Option A: Dedicated Folder

User creates "Shared with Gandalf" folder on reMarkable → moves items there → agent fetches with rmapi mget

Option B: Tag-Based

User tags documents with share-with-gandalf → agent discovers with rmapi find --tag

Option C: Starred Items

User stars items → agent fetches with rmapi find --starred

Fetch Script

# Fetch from shared folder
~/clawd/scripts/remarkable-fetch.sh
Fetch starred items
~/clawd/scripts/remarkable-fetch.sh --starred
Fetch by tag
~/clawd/scripts/remarkable-fetch.sh --tag="share-with-gandalf"

Notes

Tablet must cloud-sync before files are available

geta renders annotations into PDF (preferred for handwritten content)

Use convert (ImageMagick) to go from PDF → PNG for image processing

For text extraction, interpret the handwriting visually (vision model) rather than traditional OCR

Tool: rmapi (ddvk fork) v0.0.32

Binary: ~/bin/rmapi

Config: ~/.rmapi (device token after auth)

Sync folder: ~/clawd/remarkable-sync/

Authentication (ONE-TIME)

1. User goes to https://my.remarkable.com/connect/desktop 2. Logs in, gets 8-character code 3. Run rmapi and enter the code 4. Token saved to ~/.rmapi — future runs are automatic

Tablet must cloud-sync before files are available

geta renders annotations into PDF (preferred for handwritten content)

Use convert (ImageMagick) to go from PDF → PNG for image processing

For text extraction, interpret the handwriting visually (vision model) rather than traditional OCR