Generates grounded, long-form articles via STORM-style multi-perspective Q&A and retrieval.
npx skills add https://github.com/gbencke/dotfiles --skill storm
Please help me install this Agent Skill. Skill: storm Source: https://github.com/gbencke/dotfiles/tree/main/pi/skills/storm Command: npx skills add https://github.com/gbencke/dotfiles --skill storm
This Skill eliminates the pain of turning vague topics into deeply researched, well-structured, citation-backed long-form articles.
Ask the AI to run the STORM pipeline for the topic “How AI code agents work end-to-end,” stopping at each human checkpoint and producing a 50,000+ word final article with properly generated citations.
High-intent search queries and answers about installing and using this skill.
To generate a long-form research article with citations, provide a topic to trigger STORM-style synthesis. The pipeline creates an outline, gathers facts via multi-perspective Q&A, and drafts sections into a heavily cited final article.
Multi-perspective Q&A in deep research simulates different editorial viewpoints to interrogate a topic. This STORM-style mechanism generates questions, retrieves web data, and iteratively refines a topic outline into structured sections.
Yes, you can approve the research outline before generation. The workflow enforces human checkpoints at perspective selection, outline approval, and section review, ensuring you validate research depth before the final article drafts.
Web scraping and knowledge synthesis map citations by preserving visited sources locally. The pipeline generates deterministic citation mapping via a local references/index.json file and appends a References table to the final article.
The minimum word count for a generated deep research article is 50,000 words. The pipeline enforces this length for the final_article.md output to ensure comprehensive topic coverage and deep knowledge synthesis.
Research artifacts and scraped sources are stored locally under a .storm/<topic-slug> directory. This isolated environment contains the drafted sections, references/index.json, and the final assembled article for deterministic retrieval.
haizhouyuanInitialize topic research archives with templates, tasks, and mem0 insights.
NirajkashyapExport wiki knowledge graphs to graph.json, graph.graphml, cypher.txt, and graph.html.
ArogyaReddyImplement retrieval-augmented generation pipelines with vector stores and reranking.
rlagycksGenerate long-form articles and newsletters in a specified voice.
