iiRecord
Agentic AI Atlas · Research Specialization (Library)
page:library-researcha5c.ai
II.
Page reference

page:library-research

Reading · 2 min

Research Specialization (Library) reference

Processes for systematic research analysis, standards comparison auditing, and extraction verification, plus a flagship end-to-end research-publication pipeline that composes the scanner point-tasks below.

Pagewiki/library/research.mdOutgoing · 1Incoming · 1

Research Specialization

Processes for systematic research analysis, standards comparison auditing, and extraction verification, plus a flagship end-to-end research-publication pipeline that composes the scanner point-tasks below.

Processes

  • **research-publication-workflow.js** — Flagship end-to-end research pipeline: question framing -> plan -> parallel source gathering (composing the five scanner point-tasks below as lane playbooks) -> adversarial claim verification with re-fetched citations -> cited synthesis -> policy-gated publication. Deepest kip-librarian integration (kind research).
  • **standards-gap-audit.js** — Generic gap audit process for standards research documents. Audits research/comparison documentation against source extraction text using configurable failure pattern categories.
  • **evangelist.js** — Evangelist persona point-task: scans recent project activity (commits, PRs, docs, releases, benchmarks), filters for marketable novelties, and opens an evangelist report issue per item.
  • **novelties-scanner.js** — Novelties Scanner persona point-task: detects, analyzes, and reports on novel innovations, emerging trends, and breakthrough developments.
  • **patentable-novelties.js** — Patentable Novelties persona point-task: extends the novelties scanner with patent-potential assessment and structured invention disclosures.
  • **vendor-researcher.js** — Vendor-researcher persona point-task: discovers candidate vendors, analyses them in parallel against criteria, and produces a comparison and recommendation report.
  • **news-intelligence-pipeline.js** — End-to-end sourcing + intelligence pipeline (folded in from the former sourcing/ specialization): discover -> monitor -> dedupe -> filter signal -> per-portfolio impact assessment -> synthesize -> route alerts -> track follow-through. The former path specializations/sourcing/news-intelligence-pipeline.js remains as a header-only @deprecated alias that re-exports this file.

Usage

js
import { process } from './standards-gap-audit.js';

const result = await process({
  documents: [
    { name: 'Comparison Doc', path: '/path/to/comparison.md' },
    { name: 'Engineering Changes', path: '/path/to/changes.md' }
  ],
  extractionFile: '/path/to/extraction.txt',
  extractionDir: '/path/to/extraction/',
  // Optional: custom gap patterns and fix instructions
  gapPatterns: [...],
  fixInstructions: {...},
  domainContext: 'Steel design standard comparison'
}, ctx);