AI Engineering

Piyaz Decompose Feature turns a new capability request for an active Piyaz project into a coherent cluster of tasks. It reuses the project’s existing categories and tags, adds internal and existing-task edges, and leaves project status unchanged.

Quest Passthrough Camera Access exposes the forward-facing RGB cameras on Quest 3 and 3S in Unity. Use it to capture camera frames, read pose and intrinsics, and project pixels into world space for computer vision and machine learning.

Unity Face Tracking drives ARKit-rigged head and face blendshapes from Quest face tracking and audio-to-expression input. Use it when an FBX or mesh with the 52 ARKit shapes should animate on Quest Pro, Quest 3, or Quest 3S.

Unity Project Analyzer builds and maintains a living .agent-docs/ knowledge base for Unity projects. Use it to scan project structure, explain how a system works, or update documentation after structural changes.

Vedic Analysis interprets birth data through Vedic astrology, Nakshatra, and Vimshottari Dasha. Use it for karmic themes and life-stage readings from birth information or chart data.

Verify First requires authoritative Meta documentation checks before answering Quest, Horizon OS, or Meta XR questions. Use it when a task involves Quest-specific APIs, manifests, SDKs, or store policy so you avoid stale guidance.

Yuan unifies birth data into one input package and renders a combined destiny reading from methods that can be supported in the current environment. It is used for fate, astrology, and life-path readings with a five-year forecast window.

Ziwei Analysis structures Zi Wei Dou Shu work into deterministic chart facts and interpretive reading. Use it to design a Ziwei engine, interpret chart JSON or tables, or separate verified placement data from heuristic analysis.

Ask provides context-aware Q&A with automatic context gathering for codebase, git history, rules, docs, and other skills during development. Use it for quick questions when you need a sourced answer, not for code changes or deep research.

Ask answers quick questions about a codebase, git history, rules, docs, or other skills by gathering relevant context first. Use it for read-only questions during development, not for changes or deep research.

Clarify routes requests through shallow Q&A or deeper exploration based on ambiguity and complexity. Use it when a task needs requirements gathered before implementation, especially for complex or uncertain changes.

Code Investigate performs deep code analysis with independent Claude and Codex passes, then merges the findings. Use it to trace how a feature works, map data flow, or confirm complex behavior.