3. Data, Science & AI

Offers proven patterns and techniques for fine-tuning Large Language Models (LLMs) to enhance task-specific performance and adaptability.

Enables creation and configuration of custom sub-agents in Claude Code with tailored prompts, tools, and models for AI workflows.

Chains multiple AI model control tools in a research-to-implementation pipeline with graceful degradation for robust AI system deployment.

Manages human labeling tasks for AI training data including audience creation, job configuration, and model benchmarking with Rapidata SDK.

Verifies AI tools, resources, and prompts functionality by testing an MCP server locally.

Provides unified mathematical computation, problem solving, and explanations by routing to specialized tools for accuracy and efficiency.

Persistent memory system using file storage and knowledge graphs to maintain context across AI sessions, model-agnostic and compatible with any LLM workflow.

Defines and manages rules for coordinating multiple AI agents in complex LLM workflows to ensure efficient task execution.

EvalKit enables conversational creation, execution, and analysis of robust AI agent evaluations using the Strands Evals SDK.

Creates and manages multi-agent patterns within the Agentica AI framework for complex AI system configurations.

Performs geometric operations including shape creation, boolean set operations, and spatial measurements using the Shapely library.

Provides procedural memory for AI coding agents, enabling persistent cross-agent learning with confidence decay and anti-pattern avoidance via three-layer cognitive architecture.

Enables isolation of agent contexts to maintain state integrity in multi-agent AI systems.

Retrieves relevant past session learnings from an AI memory system to maintain context and inform current interactions.

Searches Lean 4 mathlib database via Loogle API for mathematical theorems, definitions, and proofs.

Graph-aware issue triage engine using network analysis metrics and AI agent protocols for dependency-aware planning in issue tracking.

Enables AI agents to continuously improve by capturing and applying learnings from errors, corrections, and external feedback.

Unified CLI/TUI to index and search local coding agent history from multiple AI assistants, purpose-built for AI agent consumption.

Deterministically routes mathematical user intent to exact CLI commands within a math cognitive stack.

Provides AI-powered web search, research, and reasoning capabilities using Perplexity's language model.

Provides tools for aggregating data and creating visualizations in dashboards, including score rings, charts, and wallboard views for analytics.

Calculates health metrics, longevity scores, and lifespan estimates from user health data for tracking and analysis.

Automates structured AI-driven development workflows by clarifying requirements, creating phased task plans, and executing via Ralph wrappers.

Manages scoring algorithms and weights for an 8-dimensional life balance system, enabling personalized dimension tracking and analysis.