Skip to main content
Rules let you fine-tune how FlipsideAI responds to your questions. Think of them as persistent instructions that shape every conversation—without having to repeat yourself.

​
What are Rules?

Rules are custom directives that control:
  • Personality: Formal analyst, casual expert, technical deep-dive, etc.
  • Style: Visual preferences, chart types, color schemes
  • Expertise: Domain-specific knowledge or frameworks to apply
  • Response patterns: Output format, level of detail, explanation depth
Once you set a Rule, it stays active across all your chats until you change it.

​
Managing your Rules

1

Access Rules

Navigate to My Rules in your FlipsideAI settings
2

Create or edit

  • Create new Rules from scratch - Browse Copyable Rules for community templates - Edit existing Rules to refine behavior
3

Apply Rules

  • Set default Rules that apply to all new chats - Switch Rules mid-conversation when you need different behavior

​
Rule categories

​
Personality Rules

Define the tone and expertise level of responses. 
Act as a quantitative analyst with deep DeFi expertise.
Prioritize data accuracy and statistical rigor.
Include confidence levels and caveats where appropriate.
Use precise terminology and explain complex calculations step-by-step.
Provide high-level insights suitable for non-technical stakeholders. Lead with key
findings and actionable takeaways. Use plain language and avoid jargon. Keep analyses concise—no
more than 3-4 key points per topic. ```
</Accordion>

<Accordion title="Research mode" icon="flask">
  ```text
  Approach analyses with scientific rigor.
  Always validate findings with multiple data sources.
  Explicitly state assumptions and limitations.
  Suggest follow-up questions to deepen understanding.

​
Style Rules

Control visual output and formatting. 
All charts should be optimized for Twitter:
- Use 16:9 aspect ratio
- Bold, high-contrast colors
- Large, readable fonts
- Minimal text, maximum visual impact
- Dark background with vibrant accent colors
Format all outputs for slide deck inclusion: - Clean, minimalist design - Use corporate
colors: primary #7356ff, secondary #f5f5f5 - Include clear titles and axis labels - Limit data
points to avoid visual clutter - Add source attribution to all charts ```
</Accordion>

<Accordion title="Data tables preferred" icon="table">
  ```text
  When presenting results, prioritize tables over charts.
  Use clear column headers and sort by the most relevant metric.
  Include summary statistics at the bottom of tables.
  Highlight outliers or notable values.

​
Expertise Rules

Apply domain-specific frameworks or methodologies. 
When analyzing addresses or transactions:
- Look for patterns indicating wash trading, sybil attacks, or coordinated behavior
- Flag suspicious activity (rapid token swaps, circular transfers, etc.)
- Cross-reference with known labels and tags
- Suggest additional checks or red flags to investigate
When analyzing tokens, always include: - Supply distribution (top holders, concentration
metrics) - Transfer patterns (velocity, holding time) - Liquidity depth and DEX presence - Price
correlation with market conditions - Potential risks (centralization, liquidity concerns) ```
</Accordion>

<Accordion title="Protocol health metrics" icon="heart-pulse">
  ```text
  When evaluating protocols, assess:
  - User growth and retention (DAU, MAU, cohort behavior)
  - Revenue and fee generation
  - TVL trends and composition
  - Competitive positioning
  - Key risks and sustainability concerns

​
Response pattern Rules

Shape how FlipsideAI structures its answers. 
Keep all responses brief and to the point.
Lead with the answer, then provide supporting data only if needed.
Avoid lengthy explanations unless explicitly asked.
Use bullet points instead of paragraphs.
Explain not just the but the and Define technical terms and concepts
inline. Provide context about data sources and methodology. Suggest related topics or deeper dives
the user might find interesting. ```
</Accordion>

<Accordion title="Comparative analysis" icon="scale-balanced">
  ```text
  Always structure analyses as comparisons when possible.
  Use relative metrics (percentages, ratios) alongside absolute numbers.
  Highlight outliers and explain why they differ.
  Include benchmarks or industry standards for context.

​
Example: combining Rules

You can stack multiple Rules to create sophisticated behavior: 
Personality: Technical analyst with DeFi expertise
Style: Presentation-ready charts with corporate colors
Expertise: Protocol health metrics
Response pattern: Concise, bullet-point summaries

When analyzing protocols:
1. Lead with 3 key health metrics (user growth, TVL, revenue)
2. Use clean bar charts suitable for slides
3. Keep text explanations under 50 words per point
4. Flag any sustainability concerns or risks

​
Pro tips for effective Rules

Vague Rules like “make it look nice” don’t work well. Instead:
  • Specify exact colors, fonts, or design elements
  • Define thresholds for what counts as “significant”
  • Give examples of what you want
Start with a simple Rule, run a few analyses, then refine based on results.Try asking: “Does this output match my Rule? If not, how can I improve the Rule?”
Upload example images, reference external style guides, or link to sample analyses you admire.Example: “Style charts like Dune Analytics featured dashboards: [link]”
Maintain multiple Rule configurations:
  • “Quick research” for exploratory work
  • “Client presentation” for polished outputs
  • “Internal team” for technical deep-dives
Swap between them as needed.

​
Browse community Rules

Check out the Copyable Rules library to see how other users customize their experience. Copy and adapt Rules that fit your workflow.

​
When Rules don’t apply

Rules are persistent instructions, but you can override them in any conversation:
  • “Ignore my usual Rules for this question—I just want a quick number”
  • “Follow my presentation Rules, but use a line chart instead of a bar chart”
  • “Use my technical analyst Rule, but explain it like I’m new to DeFi”

​
Next steps

Master prompting

Combine great prompts with effective Rules

Use Workflows

Apply specialized methodologies to complex analyses

Work with teams

Share Rule configurations with your team

Analysis modes

Understand how Rules interact with different modes