Orchestrates multi-phase research-to-article generation using the Stanford STORM methodology.
npx skills add https://github.com/FradSer/dotclaude --skill storm-engine
Please help me install this Agent Skill. Skill: storm-engine Source: https://github.com/FradSer/dotclaude/tree/main/storm/skills/engine Command: npx skills add https://github.com/FradSer/dotclaude --skill storm-engine
Writing well-grounded, citation-backed long-form articles requires coordinated research, outlining, drafting, and polishing; this Skill encodes the Stanford STORM methodology as a resumable, stage-gated pipeline so every /storm:* skill produces verifiable articles instead of hallucinated content.
Ask Claude to run the storm research phase on your topic, for example: generate a STORM article about retrieval-augmented generation and save it to my docs folder.
High-intent search queries and answers about installing and using this skill.
Run the storm pipeline phases in order: research, outline, write, then polish. The research phase collects sources via MCP search tools or WebSearch, and every inline [n] citation in the final article maps to an entry in sources.json.
STORM is a Stanford method (NAACL'24/EMNLP'24) that discovers multiple personas, simulates question-answer conversations grounded in web retrieval, then refines an outline and writes each section with citations. This Skill ports that pipeline into Claude-native phases.
Yes, each phase is independently resumable. The engine checks whether the phase's output artifact exists and is non-empty, skipping completed phases and loading their artifacts unless you pass --force.
Yes, it falls back to the built-in WebSearch and WebFetch tools when no MCP search tool is available. You can also pass --docs to ground research in local files instead of or alongside web results.
Skipping research would produce an article from parametric memory alone, which hallucinates citations and loses multi-perspective grounding. The engine requires research/sources.json with at least one entry before outlining or writing.
gbenckeGenerates grounded, long-form articles via STORM-style multi-perspective Q&A and retrieval.
HeadyAIPublish and synchronize content across Heady CMS sites via MCP tools.
dobachiGuide academic paper structure, literature reviews, and citation styles.
jamesacarrSearch the web via the Tavily API for content snippets and metadata.
