Deep Research Skill

Claude skill that runs an 8-phase research pipeline with source credibility scoring and automated validation. Doesn’t just paste a top search result — works through identification, cross-referencing, scoring, synthesis.

Pipeline (per source)

1–8. Eight discrete phases (specifics not detailed in source). Each phase produces an intermediate artifact; the final output is a synthesized report with source credibility scores attached.

Type and platform

  • vs gpt-researcher — GPT Researcher is a standalone OSS repo (model-agnostic); Deep Research Skill is Claude Code-native.
  • vs the Anthropic Deep Research product — different surface; this is a community skill, not a hosted service.

Hands-on notes

(empty)

Sources

See Also