AI Engineering

Auto Paper Improvement Loop reviews a compiled paper with GPT-5.4, applies writing and consistency fixes, and recompiles it for up to two rounds. Use it to iteratively polish a generated paper after drafting.

Auto Review Loop runs repeated review, fix, and re-review cycles until an external reviewer gives a passing assessment or the round limit is reached. Use it for autonomous iterative improvement of research or code.

Dr Claw helps inspect projects and Claude/Cursor/Codex/Gemini sessions, check TaskMaster progress, and generate OpenClaw-ready JSON reports. Use it for workspace lookup, status checks, and event-driven mobile updates.

Experiment Bridge implements experiment code from an experiment plan, deploys it to available GPU resources, and collects initial results. Use it when a plan is ready to execute or you need to bridge idea discovery into the auto-review loop.

Idea Discovery Pipeline orchestrates literature review, idea generation, novelty checking, and research review to turn a broad direction into validated ideas. Use it when you want a full workflow for finding and testing research directions.

Novelty Check verifies whether a proposed research method or idea has already appeared in recent literature. Use it before implementation when you need a blunt assessment of overlap and likely prior work.

Research Idea Creator generates, validates, and ranks publishable research ideas from a broad topic. Use it when you need brainstormed directions, quick novelty checks, and a shortlist of ideas to test.

Ctx uses local agent history as working memory for prior decisions, attempts, and evidence. Use it to ground new work, resume investigations, or trace code and Git artifacts back to the agent session that produced them.

Codex Guide explains how to use Codex CLI, app, and IDE features, including rules, hooks, AGENTS.md, skills, subagents, and config. It is used when answering Codex usage questions and when verifying rules or approval behavior with `codex execpolicy check`.

Skill Creator generates a new SKILL.md from conversation context and task patterns. Use it when you want to turn a repeatable workflow into a reusable Claude Code skill with instructions, examples, and optional templates.

Guizang Material Illustration generates labeled explanatory illustrations, chart beautifications, and data-first editorial images from text, screenshots, or chart data. Use it when you need a supporting image for docs, slides, posts, or reports.

Agent Infini uses the InfiniSynapse CLI for multi-turn data analysis with database and RAG context. Use it to start, continue, and inspect analysis tasks and workspace files.