Sources
Per-project reference material (PDFs, URLs, YouTube transcripts, and text) that Chaptora's AI grounds its suggestions in. Managed from the Library sidebar.
How it works#
Sources are per-project reference material the AI reads alongside your chapters. Chaptora indexes each Source and uses vector search to pull the most relevant passages into the AI's context for AI Sidebar chats, Inline AI edits, and AI Agent runs.
You manage Sources in the Library. Click the Library icon in the activity bar to open the Sources sidebar. The sidebar lists every Source in the current project, with its type icon, status, and a per-row menu.
Supported types#
| Type | What gets indexed |
|---|---|
| Extracted text content | |
| Word (.doc, .docx) | Extracted text content |
| Excel (.xls, .xlsx) | Cell text |
| Markdown (.md) | Raw text |
| URL | Scraped page text |
| YouTube URL | Transcript |
| Image (.png, .jpg, .jpeg) | Used as a direct attachment when pinned in chat |
| Text | Plain text you paste |
Add a Source#
Click the upload icon in the top right of the Sources sidebar. The Add Source dialog opens with three tabs:
File#
- Stay on the File tab.
- Drag files into the drop zone, or click to browse.
- Click Add Source.
Accepted formats: PDF, Word, Excel, Markdown, PNG, JPEG. Max 20 MB per file.
URL#
- Click the URL tab.
- Paste a website or YouTube URL.
- Click Add Source.
Chaptora scrapes the page text (or pulls the YouTube transcript) on the backend.
Text#
- Click the Text tab.
- Give the snippet a title and paste your content.
- Click Add Source.
Use Text for research notes, interview excerpts, or anything you've already written elsewhere.
Watch the status#
Each Source has a status icon next to its name:
- Pending: queued for ingestion.
- Processing: Chaptora is extracting and embedding the content.
- Ready: available for the AI to use.
- Failed: ingestion couldn't complete (e.g., a paywalled URL or a corrupt file).
The status auto-refreshes every few seconds while anything is pending or processing.
Rename, preview, or download#
- Click a ready Source to open it. URLs and YouTube open in a new tab. Files download. Text and Markdown open in a preview pane.
- Per-row ⋯ menu: choose Rename Source or Delete Source.
- Bulk actions: hover over a row and click the checkbox; once any row is selected, the header shows the count and a bulk-delete button. Click X to leave selection mode.
Use Sources in AI requests#
Sources feed the AI in two ways:
- Automatic grounding: Chaptora's vector search pulls the most relevant passages into every AI request, with no extra effort from you.
- Explicit pinning: in the AI Sidebar, open the context menu above the input and pin specific Sources. Pinned Sources stay attached for the whole conversation, so the AI treats them as required context rather than letting vector search decide.
Plan limits#
The number of Sources per project depends on your plan:
| Plan | Sources per project |
|---|---|
| Octavo | 3 |
| Quarto | 50 |
| Folio | Unlimited |
If you select more files than your plan allows, the dialog asks whether to add the first N or cancel.
Limitations#
- Max upload size is 20 MB per file.
- A URL Source captures the page text as it appeared at the time of ingestion. To update it, delete and re-add.
- Image Sources aren't part of the vector grounding. They're useful when pinned as chat attachments, not as automatic context.