Why Use Claude Code Inside Scholardo, Instead of on Its Own?
Website overview page (not a step-by-step tutorial) | For all prospective users | Explaining the value of the "context moat"
You can install Claude Code, Codex, and the like elsewhere too. So why use them inside Scholardo? In one sentence: inside Scholardo, the AI can directly "see" your entire research — your library, your notes, the passages you've read — without you copy-pasting over and over.
The Pain of Using AI on Its Own
Doing academic work in a blank AI chat box means constantly "feeding" it context:
- Want it to reference a paper? First find the PDF and paste in big chunks of text.
- Want it to continue from your notes? Copy your notes in too.
- Switch questions or start a new chat? You do all of the above again.
The AI is smart, but it doesn't know what your research looks like — every time, it starts from scratch.
What's Different When You Use It Inside Scholardo
Scholardo wires your project context to the AI through scholardo-mcp. So the agent can directly reference the things in your project to answer, without you hauling them over:
- Read your library directly: "Among the papers in my sources, which ones support this claim?" — it goes through your PDFs and answers with citations.
- Continue from your notes: "Based on the points I recorded in notes, help me draft an introduction." — it reads your notes and writes along your line of thought.
- Cite passages you've annotated: the content you highlighted / annotated while reading is available to the AI too.
That's the "context moat": wherever your research materials are, that's where the AI's context is — what it saves you isn't just a bit of copy-pasting, it turns the AI from a "generic chatbot" into a "research assistant that has read all your materials."
Shot list (demo visuals)
The goal is to visualize the fact that "the AI references your library/notes directly." Capture the Scholardo UI from the current release; use a sample project for the demo to avoid exposing real research materials.
| Asset | Content | UI source | Key callout |
|---|---|---|---|
| GIF / video 1 | Left: a sources list with a few PDFs; right: ask in the terminal "which ones support this claim," and the AI answers with filenames/citations | Scholardo | Circle the cited sources in the answer; caption: "it's reading your library" |
| GIF / video 2 | A few notes in the Notebook tab; in the terminal, have the AI "draft an introduction based on my notes," and it writes along the notes' content | Scholardo | Circle the note → answer correspondence |
| Fig 1 | Comparison diagram: left "blank AI = repeated copy-pasting" vs right "Scholardo = the AI sees your library/notes directly" | Custom diagram | Emphasize "no hauling — the context is already in the project" |
In One Sentence
Use AI on its own and you keep feeding it materials; use it inside Scholardo and the AI is already standing in your materials. That's the core reason to put Claude Code inside Scholardo.
Data Privacy Disclosure
When the AI references your library, notes, or annotations to answer, the referenced manuscript content is sent to your chosen AI provider for processing (Anthropic / OpenAI / DeepSeek, etc.). Scholardo only sends the corresponding content when you actively start a conversation and the question genuinely needs that material — it won't upload your whole project without your knowledge. If your research materials involve unpublished data, participant privacy, or confidentiality agreements, first confirm that sending them to a third-party AI service complies with your institution's requirements.