AI Engineering

Manim Community Best Practices covers Scene structure, animations, MathTex, 3D scenes, camera control, styling, and CLI usage for Manim Community Edition projects. Use it when writing or reviewing ManimCE code and render commands.

Manim Composer turns vague educational video ideas into scene-by-scene Manim plans. It researches the topic, asks clarifying questions, and outputs a scenes.md outline before any animation code is written.

Ex Persona Builder guides users through chat history analysis and questions to create a digital persona from an ex-partner’s communication style. It is used when someone wants to reconstruct tone, habits, and response patterns for later simulation.

Zhimin Liu Persona defines a character profile and response rules for roleplaying as Zhimin Liu. Use it when you want an AI to stay in character, follow the listed emotional modes, and match the user's language.

Paper Analyzer turns academic papers into detailed, shareable HTML essays with formula rendering, Mermaid diagrams, and codebase cross-checks. Use it when you want a paper explained clearly from abstract to implementation.

Paper Deck turns papers and technical content into high-fidelity slide decks. It plans the narrative first, then generates 16:9 slide images page by page and compiles them into PPTX or PDF.

Paper Method Illustrated turns a paper’s core method into visual diagrams that explain the workflow and key mechanisms. Use it when you need to plan cover, overview, or detail figures for a paper in Chinese, English, or both.

LLM Wiki Manager initializes, ingests, compiles, queries, and audits LLM-compiled wikis. Use it for knowledge-base work, inventories, idea and project portfolios, archives, sessions, and feedback.

LLM Wiki Manager ingests sources into an LLM-compiled wiki, compiles articles, and supports querying, linting, auditing, archiving, and session capture. Use it for wiki workflows, idea curation, portfolios, datasets, and related knowledge-base management.

LLM Wiki Manager manages an LLM-compiled knowledge base in Codex, including ingesting sources, compiling articles, querying, linting, auditing, archiving, and session capture. Use it for wiki workflows, topic portfolios, datasets, ideas, and research plans.

Agent Debug captures a stuck, looping, or drifting agent run and identifies the likely failure pattern. Use it when an agent is burning tokens, repeating tools, or failing on recoverable assumptions and you need the smallest safe recovery step.

Agent Ops Review helps managers inspect AI-agent progress, blockers, quality gates, and throughput levers. Use it when you need a structured status view of agent work and its next actions.