# Instapaper Import

> Import reading notes from Instapaper exports into the slipbox. Use when the user pastes an Instapaper highlight export (article title + notes).

- Skill: `jrswab/instapaper-import` (Agent Skill)
- Install (CLI): `npx skillmds@latest add jrswab/instapaper-import`
- Raw SKILL.md: https://api.skillmd.com/api/skills/jrswab/instapaper-import/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: jrswab (https://skillmd.com/u/jrswab)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/jrswab/instapaper-import

---


# Instapaper Import

Parse Instapaper highlight exports and create slipbox entries for the user's own notes.

## Input Format

```
# [[Article Title](url)]
> Original highlight from article (SKIP)
User's note about the highlight (IMPORT)
> Another highlight (SKIP)
Another user note (IMPORT)
```

- `> lines` = original article highlights → **skip**
- Plain text lines = user's own notes → **import** as separate entries
- Empty lines → skip

**Header fields:**
- Title + URL from `# [[Title](url)]`; URL may be `instapaper-private://...` or a regular URL
- Source type: `article`
- Author: `null` (Instapaper doesn't include author)

## Workflow

1. **Parse header** → extract article title and URL. **Done when** title and URL are both captured.
2. **Extract user notes** → collect every plain-text line (not starting with `>`, not empty). **Done when** every non-`>`, non-empty line in the input is accounted for.
3. **Precheck** → show user: article title, note count, ask for confirmation. **Done when** user confirms (or aborts).
4. **On confirmation** → for each note, invoke slipbot:
   - Type: note (`-` prefix)
   - Source: `~ article, {title}`
   - Let slipbot handle: filename, tags, links, graph update
5. **Report** → count of notes created. **Done when** the reported count matches the extracted count.

## Example

**Input:**
```
# [[How to Learn Faster](https://example.com/article)]
> Get feedback more often
To learn faster we need faster feedback loops.
> Latent learning occurs without reinforcement
Testing yourself proactively speeds up learning.
```

**One resulting slipbot call:**
```
- To learn faster we need faster feedback loops. ~ article, How to Learn Faster
```

## Edge Cases

- **No user notes** (only `>` lines): Report "no notes to import"
- **Multi-line user notes**: Treat each paragraph as a separate note
- **Title with special chars**: Preserve as-is for source metadata