
AI Context Engineering Suite

AI Context Engineering Suite
ContextIQ is an AI context engineering suite designed for engineers working with large language models (LLMs) and AI applications. The platform provides a set of visual tools that help developers reason about and manage the context used by AI systems. It is developed and maintained by Trango Compute Inc. and is available at contextiq.trango-compute.com. The core tools include a RAG Chunk Inspector, which shows how documents are split into chunks using different strategies (tiktoken, sentence, paragraph) with a live LLM context preview; an Agent Workflow Visualizer that accepts a GitHub URL to instantly render agent graphs for frameworks such as LangGraph, CrewAI, AutoGen, Google ADK, and OpenAI Agents SDK; and an Agent Trace Inspector that parses OTLP JSON traces from platforms like LangSmith, Langfuse, or OpenTelemetry to display full agent graphs with per-node token attribution for supported frameworks. Additional features include a Memory Architecture Visualizer for designing agent memory layers as directed acyclic graphs (DAGs) with episodic, semantic, and procedural memory mapping; a Token Inspector for comparing token counts and costs across models like GPT-4o, Claude, Gemini, and over 20 others; a HyDE Visualizer for comparing direct query versus hypothetical document embedding retrieval using all-MiniLM-L6-v2 cosine scores across corpus chunks; and an OIDC Inspector for scanning domains to detect OpenID Connect configurations from providers such as Google, Microsoft, Auth0, and Okta across 11 subdomains. Typical use cases for ContextIQ include debugging and optimizing retrieval-augmented generation (RAG) pipelines, visualizing agent workflows and trace data during development, designing and validating memory architectures for multi-turn agents, estimating token usage and costs before committing to a model, assessing the effectiveness of hypothetical document embeddings, and auditing identity provider configurations in web applications. The workflow generally involves pasting a document, URL, or trace data into the appropriate tool, interacting with the visual output to understand context handling, and using the insights to refine the AI system. All tools are available for free without requiring a credit card. A paid upgrade at $5 per month enables export of results as PNG, SVG, or PDF files and provides shared links. No specific installation or technical environment beyond a modern web browser is required to use the online suite.