SuperAPI AI Directory

Skills

Find an Agent Skill for your task. Read the guide and copy the install command.

Updated 11-07 00:13

Writing & research

14,945 results

amazon-keyword-research

Amazon seller keyword research and market opportunity analysis. Get autocomplete suggestions (long-tail keywords), analyze the competitive landscape, and evaluate the market opportunity for any keyword across 12 Amazon marketplaces (US/UK/Germany/France/Italy/Spain/Japan/Canada/Australia/India/Mexico/Brazil). No API key required. When users mention Amazon product research, finding products to sell on Amazon, Amazon keyword ideas, niche analysis, Amazon competition...

Writing & research1 installs

stem-explainer-video

STEM science-popularization video generation skill. For explainer videos on science, technology, engineering, and mathematics topics: first use AI to generate a structurally strict Chinese schematic diagram, then use MiniMax H3 to drive animation with that image as the first frame, and finally use Remotion to deterministically overlay subtitles, terminology cards, and annotations, outputting a vertical or horizontal final video. Trigger words: science popularization video, explainer video, schematic animation, knowledge explainer video, STEM…

Writing & research1 installs

doc-writing

Write, revise, or translate project documentation in Chinese and English (README, user guide, design doc / RFC, module description). First get the facts, the readers, and the trade-offs from the maintainer, then write this material into documentation that is easy for readers to read: decide what matters first, start every sentence from something the reader already has, and follow abstract statements with consequences and examples. Use for new writing, rewriting, translation, and removing the AI flavor from documentation; not for blog posts, marketing copy, or generating API docs from source code…

Writing & research1 installs

light-consistency

Unified style and consistency maintenance. Maintain consistent terminology, visual style, logical threads, and innovation-point phrasing across papers, PPTs, charts, code, and project documents (always active, effective in the background for all tasks). Avoid the same project having inconsistent statements across different materials, non-unified metric names, chaotic chart styles, contradictory innovation points, changing method names, non-unified dataset names, inconsistent logic between paper and PPT, and inconsistency between software copyright and system functions.

Writing & research1 installs

light-idea-critique

Strictly judge by top-journal/top-conference reviewer standards whether the idea truly makes a breakthrough, or is merely a conventional combination, repackaging, concept stacking, or lacking theoretical depth or experimental support. Must be used when the user asks "is this idea viable/good enough/innovative enough/poke holes in it for me," or after m03 produces an idea. First blind then explicit criteria-setting, eight-dimension weighted scoring, five-perspective adversarial review, anti-sycophancy hard protocol, giving a verdict + Revision Roadmap, guiding back to…

Writing & research1 installs

light-idea-generation

Propose research ideas with potential, differentiation, and highlights based on the actual project situation. Use when the user needs innovation points, research directions, or proposal plans, or asks "what can be done with this direction/data." Combine existing foundations, data conditions, technical capabilities, time cycles, and publication/competition/project goals to explain why it is worth doing, the innovation points, the advantages over existing methods, the specific problems it can solve, and the suitable forms of output and submission levels.

Writing & research1 installs

light-literature-search

Powerfully search and standardize the organization of research materials. Use when the user needs to investigate a research direction, find literature reviews, understand what work exists in a field, or gather Chinese and English papers/patents/standards/policies/datasets/open-source projects/competition solutions/industry reports/technical blogs/official documentation. Do not merely list results; instead filter, deduplicate, and classify, judge credibility and importance, and organize into literature tables, research lineages, representative methods, pros and cons comparisons, and reusable resources.

Writing & research1 installs

light-paper-polishing

Polish papers module by module — language polishing, logical restructuring, structural optimization, strengthening novelty, reinforcing argumentation, abstract refinement, introduction enhancement, deepening experimental analysis, and conclusion improvement. Use when the user needs to revise a paper or feels that a passage is not fluent / weak in logic / has an insufficiently prominent novelty. From the reviewer's perspective, identify places with weak logic, insufficient argumentation, unprofessional expression, vulnerability to challenge, missing citations, or needing reorganization, and provide concrete revision plans and revised text.

Writing & research1 installs

light-research-ethics

Research ethics, academic norms, and risk review. In all research tasks, by default check for risks such as academic misconduct, data fabrication, image reuse, nonstandard citation, privacy leakage, copyright, exaggerated conclusions, overpackaging, paper ghostwriting, patent ownership, and untruthful software copyright materials (always active, effective in the background for all tasks). Conduct compliance review of papers, projects, software copyright, patents, and competition materials to ensure content is truthful, standardized, explainable, and traceable.

Writing & research1 installs

light-research-plan

For an idea that has been confirmed feasible, formulate an extremely detailed research plan, experimental design, and execution plan. Use when the user's idea has passed the m04 review and needs to be broken down into a complete, executable, reproducible scientific research workflow. Cover research objectives, technical approach, data pipeline, experiments/ablation/sensitivity/robustness/significance testing, timeline, risks and alternatives, and ensure reproducibility throughout.

Writing & research1 installs