Multi-agent research report generator. STORM-inspired editorial workflow with parallel section research, claims-verified review loops, and five report types (basic, detailed, deep, outline, resource). Supports web, local, and hybrid source modes. Integrates with cogni-claims for source verification.
npx claudepluginhub cogni-work/insight-wave --plugin cogni-researchExtract verifiable claims from a report draft for downstream verification via cogni-claims.
Recursive tree exploration for deep research mode — single branch, multi-query internal search.
Research a single sub-question from local files (PDF, DOCX, TXT, MD, CSV) instead of web search.
Evaluate report drafts against structural review criteria and claims verification data.
Incorporate reviewer feedback and claims deviation data into a revised draft.
Perform parallel web research for a single sub-question, creating context and source entities.
Rank, filter, and annotate research sources by quality, relevance, and diversity.
Compile aggregated research context and source entities into a report with inline citations.
Generate a multi-agent research report using parallel web research with structural review. Three modes: basic (fast single-pass), detailed (multi-section with outline), deep (recursive tree exploration). Claims verification runs separately via verify-report. Supports configurable writing tones, auto/manual researcher roles, source URL pre-fetch, domain-restricted search, custom sub-question counts, and local document research. Three source modes: web (default), local (analyze user's documents), hybrid (web + documents). Use when the user asks to "research report", "investigate", "deep research", "write a report", "gpt-researcher", "multi-agent research", "analyze these documents", "research from my files", or requests comprehensive topic analysis with citations. Also use when the user wants to "resume research", "continue research report", "pick up the research", "finish the report", "what happened to my report", or resume an interrupted research run.
Verify claims in a research report against cited sources using cogni-claims. Works on cogni-research project directories (auto-detects draft, sources, and existing claims) or on any standalone markdown report with inline citations. Use when the user asks to "verify report", "verify claims", "check sources", "fact-check the report", "verify the research", "run claims verification", "check the citations", or wants to re-verify a report after editing. Also use when the user says "verify" after a research-report run completes, or when research-report's Phase 6 summary recommends running verify-report.
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Complete collection of battle-tested Claude Code configs from an Anthropic hackathon winner - agents, skills, hooks, rules, and legacy command shims evolved over 10+ months of intensive daily use
Upstash Context7 MCP server for up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your LLM context.
Comprehensive startup business analysis with market sizing (TAM/SAM/SOM), financial modeling, team planning, and strategic research
AI-powered development tools for code review, research, design, and workflow automation.