SubnetAPI's profile picture. Building a tool that aggregates and tests Bittensor Subnet APIs in one place to reduce friction for Web2 developers.

SubnetAPI

@SubnetAPI

Building a tool that aggregates and tests Bittensor Subnet APIs in one place to reduce friction for Web2 developers.

It's been a little bit. Life has been very full. I've made some slow progress. Currently in the throws of testing the app and confirming the data being displayed matches what has been accumulated. Sorry for the inconsistency. I'll have more to share soon.


I have a rough draft @CryptoBuffalo

SubnetAPI's tweet image. I have a rough draft @CryptoBuffalo

Scraped 25 subnets with API information. The database is being built.


Spent the last several days working Broadcast and editing video for a production. It's been a welcome break to help me digest. Realized trying to set up a team of 41 agents is VASTLY complex. Have reduced the team to 7.


I think reframing this project's goals is a safe and healthy implementation. The loftiness of creating a piece of software appears to be part of the AI Psychosis problem. So, while I aim to produce a piece of software, the goal is to learn about Agents.


Read a fascinating post about AI Psychosis by Dave Shapiro yesterday. Definitely a cautionary explanation of an emerging headspace. Got to do my best to remain grounded. Also realize the need for verifying the research I've collected and will be having agents reference.


So true... This is the 3rd attempt at restarting this idea. At a certain point the linter errors are 10 high and the agent just disregards them as preexisting issues that it doesn't need to address. At least using Cursor. It'll be interesting using subagents.

SubnetAPI's tweet image. So true... This is the 3rd attempt at restarting this idea. At a certain point the linter errors are 10 high and the agent just disregards them as preexisting issues that it doesn't need to address. At least using Cursor. It'll be interesting using subagents.

Day Job calls today. Setting up for Preseason Football with CryptoBuffalo @TaoCity. Will do my best to squeeze some progress and updates in.


Set up this folder structure to compile research. The standards documents are critical for the Subagent Team I've collaborated on Claude with. Getting close to building them out. But I've got to set up for interviews with my wife today. Will update later.

SubnetAPI's tweet image. Set up this folder structure to compile research. The standards documents are critical for the Subagent Team I've collaborated on Claude with. Getting close to building them out. But I've got to set up for interviews with my wife today. Will update later.
SubnetAPI's tweet image. Set up this folder structure to compile research. The standards documents are critical for the Subagent Team I've collaborated on Claude with. Getting close to building them out. But I've got to set up for interviews with my wife today. Will update later.

This is really a GIANT experiment. Can a Vibe-Coder set up an agentic workflow that executes near seamlessly because I provide enough foundational research for each agent to reference; have I simplified, compartmentalized, and orchestrated this optimally? I guess we'll find out.


Working with 10 instances of Claude in parallel to research and generate some comprehensive coding standards documents.


Crypto Buffalo has a @TaoCity discord server that is free to join and is going to be the primary content channel that I work with while I build this out, aside from X.


United States Trends

Loading...

Something went wrong.


Something went wrong.