3. Data, Science & AI

Refines prompts for GPT models using OpenAI best practices to enhance output quality and task success.

Refines user prompts to enhance AI model responses, improving output quality and relevance for better results.

CLI tool for querying xAI's Responses API with support for multi-turn, streaming, and job-file polling modes in developer workflows.

Enhances ZenML MLOps workflows with recommended improvements including experiment tracking, metadata logging, and model governance features.

Exports ML-Agents models to ONNX for Unity deployment and shares them on HuggingFace Hub for collaborative access and reuse.

Optimizes ML-Agents training speed, memory consumption, and inference performance for accelerated deployment cycles.

Trains Unity ML-Agents reinforcement learning agents using PPO, SAC, or MA-POCA algorithms for AI-driven simulation environments.

Diagnoses and resolves common training issues in ML-Agents, including convergence, connection failures, and performance bottlenecks.

Provides database management utilities for table creation, migrations, and secure queries, including encryption and prepared statements for the cf7_api_logs table.

Write and validate Databricks SQL queries for game analytics, including user behavior metrics and schema exploration.

Accesses real-time and historical electricity market data including LMP, load, and fuel mix from major grid operators like ERCOT, CAISO, and PJM.

Transforms simple prompts into context-aware, comprehensive versions for optimized AI model interactions, including Context7 prompt engineering guides.

Enables autonomous AI agent coordination for complex task execution and decision-making in LLM-driven applications.

Queries DX metrics database to analyze developer productivity, team performance, and engineering KPIs including DORA metrics.

Provides guidance on selecting and applying mathematical tools within the cognitive stack for problem-solving.

Provides text-to-speech functionality using ElevenLabs API with a macOS 'say' command-like user experience.

Creates database indexes during data insertion to optimize query performance and data retrieval.

Generates unambiguous, reliable prompts for Agentica and REPL agents to ensure precise LLM instruction execution.

Retrieves past reasoning steps to inform current decisions and approaches in AI agent workflows.

Defines standardized interfaces for parallel AI agents to ensure interoperability and consistent behavior in multi-agent systems.

Routes user queries to the appropriate level of summary analysis (tldr) based on query complexity and intent.

Enables formal theorem proving with integrated research, testing, and verification phases for mathematical logic.

Validates Terminal User Interface output via screenshot capture and LLM-based semantic analysis for visual and text validation.

Executes GPT-5 Pro for extended second opinions via background processes, with session-based result retrieval.