deep-research

v2.10.3

Disciplined, multi-source meta-research for high-stakes questions — the heavyweight alternative to the fast research router. Runs a 9-phase pipeline (reframe into falsifiable hypotheses → genre/blocks → plan → capability discovery → parallel sub-agent fan-out → score & triangulate → synthesize + adversarial pass → verify → refresh targets), triangulates every thesis against >=3 independent, differently-typed sources, saves each source to its own file with verbatim quotes, and never fabricates a citation. Output is an auditable, reusable folder with a delta-update refresh protocol. Use when a wrong answer is expensive: strategy, comparing N options, hypothesis validation, mapping a field.

Claude Code1 skill

by Alireza RezvaniMIT24.2kupdated 10 hours ago

Source

$ git clone https://github.com/alirezarezvani/claude-skills

Clone the source, then follow the repository's marketplace instructions for your runtime. The plugin root is research/deep-research/ inside the repository.

Layout

research/deep-research/

├── .claude-plugin/plugin.json
└── skills/deep-research/SKILL.md

Skills1

deep-researchskills/deep-research/SKILL.md

Run a disciplined, multi-source research investigation for a high-stakes question or decision — fan-out web search across many channels, parallel sub-agents, source triangulation (each claim backed by ≥3 independent sources), an adversarial review pass, and every source saved to its own file with verbatim quotes for reuse. Use when a low-quality answer is expensive: strategy work, comparing N products/methods/markets, validating a hypothesis with external data, or mapping how a field works. NOT for quick fact-checks (answer directly), structured 12-dimension competitor scoring (use competitive-teardown), or fast topic overviews where the decision risk is low (use the research router instead).

Manifests1

research/deep-research/.claude-plugin/plugin.json

{
  "name": "deep-research",
  "description": "Disciplined, multi-source meta-research for high-stakes questions — the heavyweight alternative to the fast research router. Runs a 9-phase pipeline (reframe into falsifiable hypotheses → genre/blocks → plan → capability discovery → parallel sub-agent fan-out → score & triangulate → synthesize + adversarial pass → verify → refresh targets), triangulates every thesis against >=3 independent, differently-typed sources, saves each source to its own file with verbatim quotes, and never fabricates a citation. Output is an auditable, reusable folder with a delta-update refresh protocol. Use when a wrong answer is expensive: strategy, comparing N options, hypothesis validation, mapping a field.",
  "version": "2.10.3",
  "author": {
    "name": "Alireza Rezvani",
    "url": "https://alirezarezvani.com"
  },
  "homepage": "https://github.com/alirezarezvani/claude-skills/tree/main/research/deep-research",
  "repository": "https://github.com/alirezarezvani/claude-skills",
  "license": "MIT",
  "skills": [
    "./skills/deep-research"
  ],
  "attribution": {
    "contributed_by": "Socialpranker (https://github.com/Socialpranker), PR #851",
    "upstream": "https://github.com/Socialpranker/claude-deep-research — the continuously-maintained source catalog (29 channels, 460+ statistical sources, 39 validated APIs, 103 report blocks) referenced from references/full-catalog.md",
    "derivation_note": "Methodology SKILL.md + upstream-catalog pointer contributed via PR #851. Hardened for this repo: packaged as a full research plugin (plugin.json + marketplace entry + cs-deep-research agent + /cs:deep-research command + README) to match the research/ per-skill-plugin convention, and dropped a dangling claude-api cross-reference."
  }
}