OpenClaw for Researchers: Deep Work with AI

# Research Assistant

## Role
You are my research assistant. Your job is to help me understand, synthesise, and build on information. You are not a search engine — assume I've already found the source and hand you the content to process.

## Behaviour
- When I share a paper or article, automatically extract: key claim, methodology (if applicable), findings, limitations, and how it relates to my research questions
- Track contradictions across sources I've shared — flag when new information conflicts with something I gave you previously
- When summarising, be accurate over being concise. I'd rather a long accurate summary than a short one that loses nuance
- Cite sources by the short title or author+year I use

## Research Context
- Current research question: [your main research question]
- Related projects: [list ongoing research]
- Key databases I use: [Google Scholar, PubMed, arXiv, etc.]
- Note-taking system: [Notion/Obsidian/plain text]

## Memory Conventions
- Store paper summaries as: "paper: [author year] - [short title]"
- Store my conclusions as: "conclusion: [topic] - [my position]"
- Store open questions as: "open question: [question]"

Here is the abstract and key sections of a paper:
[paste content]

Extract and store:
1. Citation: [Author, Year, Title]
2. Core claim (1 sentence)
3. Methodology (2 sentences)
4. Key findings (bullet list)
5. Limitations (bullet list)
6. Relevance to my research question on [your topic]

File it under my research on [topic area].

Summarise this article for my research on [topic]:
[paste or URL if you have web access configured]

I need:
- The main argument
- Evidence or data cited
- Any claims I should verify
- Whether this supports or challenges the [position/finding] I noted earlier

I'll paste a book chapter in sections. After each section, briefly note:
- Main point
- Any specific data or claims worth flagging
- How it connects to my existing notes on [topic]

After all sections, synthesise the chapter's contribution to my understanding of [topic].

Based on the sources I've shared with you on [topic], synthesise:
1. What is the consensus position (if any)?
2. Where is there genuine disagreement?
3. What evidence is strongest?
4. What questions remain unanswered?
5. What should I read next based on the gaps?

From the sources I've stored on [topic], create a structured map showing:
- Core theoretical frameworks
- Key empirical findings
- Major ongoing debates
- Prominent authors/groups and their positions

Format as a structured outline.

I'm considering the hypothesis: [your hypothesis]

Based on what you know about my research on [topic], evaluate:
1. What evidence I've already collected supports this
2. What evidence contradicts it
3. What I'd need to confirm or refute it
4. Known studies or arguments I should check

Add a conclusion to my research notes:
Topic: [topic]
My position: [your current view]
Based on: [brief evidence reference]
Confidence: [high/medium/low]
Date: [today]

Add to my open research questions:
Question: [the question]
Why it matters: [brief reason]
Possible approaches: [any initial thoughts]

Review my research progress this week:
- What new sources did we process?
- What conclusions did I reach or update?
- What open questions came up?
- What should be my focus next week based on the gaps?

integrations:
  notion:
    token: "secret_your-notion-token"
    research_db_id: "your-database-id"

Save this paper summary to my Notion research database:
[paste the summary]
Tags: [topic tags]

integrations:
  filesystem:
    allowed_paths:
      - "/Users/yourname/Obsidian/Research"

Create a new note in my Obsidian research vault for this paper:
[paper summary]
Use the template: # Author Year - Title / ## Summary / ## Key Points / ## Related

I'm interviewing [role/type of person] about [topic].
Based on my research notes on this topic, suggest:
1. 8 core interview questions
2. 4 follow-up probes for likely responses
3. Topics my existing research suggests I may have blind spots on

Here is an interview transcript. Code it for themes related to my research on [topic].
Identify:
- Direct quotes relevant to [theme 1], [theme 2], [theme 3]
- Unexpected themes that emerged
- Contradictions with other interviews I've shared

[paste transcript]