Welcome to the repo for

The Center for Llamas who want to Curve good

and do governance stuff good too

---

To run

In the root directory enter the following in your terminal.

python wsgi.py

---

Overview

Curve has a complex governance ecosystem.

CRV holders lock in the locker to vote on Gauge Votes on chain (among other things).

Some CRV is locked in a wrapper contract such as Convex's cvxCRV and delegates its votes to vlCVX votes via snapshot.

This dashboard hopes to clearly present the activity of Curve Governance and expand into its various wrapped markets to complete the picture.

---

To Do

[] Snapshot votes prior to December 22 arrive in a distinct format. * Need to update to resolve: - Choices: [x, y, z] - Vote_Option: [x, y, z] * Currently handles Vote_Option as dict. > Flipside will not be retroactively updating this so need to pull prior data from Snapshot via rate limited process.

[X] Build out Snapshot Show page for specific views.

[] Build out StakeDAO Bounty index/show

[] Build out Votium Bounty index/show

[] Build out StakeDAO sdCRV index/show

[] Build out meta aggregation of all above.

[] Fix CRV Locked Chart bar chart to include withdraw and deposit info.

[] Fix Gauge Votes bar chart its not right for some reason.

[] Map Gauge to Pool to Name to Symbol

---

Points of Frustration

• Convex doesn't have a registry of what gauge addresses their vote choices actually vote to.

  • Convex uses swap address (sometimes token) instead of gauge address, so need matching, but helpful for future alt tracking liquidity.

  • it is hard to search an address by prefix.

  • to this effort I have:

    • queried all factory deployed pools
    • harvested the main contracts
    • queried all v2 deployed pools

  • I have matched 113 / 122 gauges to convex choices

    • 9 remaining.
    • several are cross chain, not ins cope. Many are USDC pairs and v2's.

  • Some votes do not

• Some curve tokens are just random contracts deployed by random EOAs like Frax USDC.
  • This

---

Architecture

Flask based website powered by Pandas Dataframes following a Model View Controler structure.

Model

Each folder has models.py which reads data from a CSV and forms dataframes of data.

there is no database at present and csv fed dataframes seems to be plenty effective

View

Each folder has a /templates folder which contains jinja2 template files which are hybrid html/python files. These define how to display information.

Controller

Each folder has routes.py which defines individual pages and any data handling between data source and the display. Here we define what info populates charts, tables, etc

The core application is defined in __init__.py in the /app folder

Data

At present data is simply CSV's a title appendage.

There is a jupyter notebook load_data.ipynb which taps calls Flipside by API to update these data sources.

To apply changes restart the server to precompile new dataframes.

---

General Dev Flow

1. create component folder
2. create models.py
3. create flipside query
4. create local dataframe in models.py which generates all desired views
5. create routes.py
6. assemble out the info for each view
7. create templates
8. create a jinja2 template for each view.
9. update __init__.py

• to import blueprint from routes.py
• to register the new blueprint

---

Reference Links

• Currently using Boostrap Themes.

  • Current theme: https://bootswatch.com/minty/

• Tables are reliant on a Javascript library Grid.js

  • Eventually should make server side sorting
  • https://gridjs.io/docs/examples/server-side-sort

• Charts are reliant on Flask-Plotly

  • https://github.com/alanjones2/Flask-Plotly
    • example
    • https://github.com/alanjones2/Flask-Plotly/blob/main/plotly/app2.py

• URL structure is managed by Flask Blueprints

  • https://flask.palletsprojects.com/en/1.1.x/blueprints/