Math, Inc. avatar

Math, Inc.

Official

@math-inc · United States of America

0Followers
|
10Public Repos
|
101Published Skills

A new company dedicated to autoformalization and the creation of verified superintelligence.

Skills Distribution
DomainAI Models & ...Distributed Model .. (40%)Inference & Quanti.. (30%)Agentic Orchestrat.. (20%)Data Engineering (10%)

Agent Skills by Math, Inc.

Showing 101 vetted skills indexed across 1 GitHub repositories.

math-incmath-inc
1.3k

powerpoint

Create, edit, and inspect PowerPoint .pptx presentations via XML manipulation and pptxgenjs.

Official
Advanced
math-incmath-inc
1.2k

qmd

Index and search private notes locally with hybrid retrieval.

Official
Advanced
math-incmath-inc
1.2k

oss-forensics

Automates OSS forensics investigations across GitHub repositories using git data and APIs.

Official
Advanced
math-incmath-inc
1.2k

1password

Automate secret retrieval and injection using the 1Password CLI.

Official
Intermediate
math-incmath-inc
1.2k

neuroskill-bci

Connect Gauss to a local NeuroSkill server to read real-time EEG/BCI metrics.

Official
Advanced
math-incmath-inc
1.2k

solana

Fetch live Solana on-chain data and USD pricing via RPC and CoinGecko.

Official
Advanced
math-incmath-inc
1.2k

blender-mcp

Control a running Blender instance remotely via TCP socket JSON on port 9876.

Official
Advanced
math-incmath-inc
1.2k

openclaw-migration

Migrate OpenClaw customization files into Gauss Agent with a structured report.

Official
Advanced
math-incmath-inc
1.2k

agentmail

Automate agent-level email communication with a dedicated AgentMail inbox.

Official
Intermediate
math-incmath-inc
1.2k

blackbox

Delegate coding tasks to a multi-model AI agent via the Blackbox CLI.

Official
Intermediate
math-incmath-inc
1.2k

telephony

Provision and manage Twilio numbers for SMS, MMS, and voice calls.

Official
Advanced
math-incmath-inc
1.2k

dogfood

Automate exploratory QA testing of web applications with structured bug reports.

Official
Advanced
math-incmath-inc
1.2k

apple-reminders

Manage Apple Reminders from the terminal using remindctl commands.

Official
Intermediate
math-incmath-inc
1.2k

imessage

Send and receive iMessages and SMS on macOS via a CLI tool.

Official
Intermediate
math-incmath-inc
1.2k

findmy

Locate Apple devices and AirTags via Find My using AppleScript and screen capture.

Official
Intermediate
math-incmath-inc
1.2k

apple-notes

Manage Apple Notes via the memo CLI on macOS.

Official
Basic
math-incmath-inc
1.2k

parallel-cli

Run asynchronous vendor-based web search and research tasks with JSON-first outputs.

Official
Advanced
math-incmath-inc
1.2k

blogwatcher

Monitor blog updates and track read status via the blogwatcher CLI.

Official
Intermediate
math-incmath-inc
1.2k

polymarket

Fetch and aggregate Polymarket market data from public REST APIs.

Official
Intermediate
math-incmath-inc
1.2k

duckduckgo-search

Search DuckDuckGo for text, news, images, and videos without API keys.

Official
Intermediate
math-incmath-inc
1.2k

ml-paper-writing

Write publication-ready ML papers with conference templates and citation verification.

Official
Advanced
math-incmath-inc
1.2k

arxiv

Query arXiv by keyword, author, category, or ID via its REST API.

Official
Intermediate
math-incmath-inc
1.2k

domain-intel

Identify subdomains, certificates, and ownership signals from DNS, SSL, and WHOIS records.

Official
Intermediate
math-incmath-inc
1.2k

minecraft-modpack-server

Automate modded Minecraft server setup from CurseForge/Modrinth pack zips.

Official
Advanced

Frequently Asked Questions About Math, Inc.

FAQPage Schema
What specific ML training tasks are supported by these capabilities?

The registry supports end-to-end training including SFT, DPO, PPO/GRPO, and reward modeling. It enables distributed pretraining with FSDP2, memory-efficient LoRA/QLoRA fine-tuning, and sparse autoencoder training for transformer activation analysis.

Which hardware environments are compatible with these inference methods?

Inference is supported across diverse hardware, including NVIDIA GPUs via TensorRT and vLLM, as well as CPU-based execution using GGUF quantization and llama-cpp for non-NVIDIA environments.

How are autonomous development tasks managed within the Gauss environment?

Development tasks are managed through subagent-driven coordination, where individual agents handle specific implementation, code review, and debugging phases. These processes utilize structured markdown planning and Pydantic-validated output constraints to ensure code quality.