Research And Documentation Workflow

User Intent

Capture generic pages, docs sites, PDFs, and research prompts into structured records with citations and reusable source evidence.

Required Inputs

• URL, domain, sitemap URL, docs URL, or research prompt.
• Optional max pages, depth, and include/exclude patterns.
• Execution mode, default auto.
• Working directory.

Run States

• Setup: target and scope controls.
• Live Execution: discovery, crawl, parse, extraction, and artifact writing.
• Results Review: records are grouped by page, section, document, or answer.

Result Tabs

• Overview: scope, page count, summary, and gaps.
• Browser: captured page or document evidence.
• Records: research/docs records.
• Sources: pages, PDFs, sitemaps, markdown.
• Blocked: inaccessible pages.
• Artifacts: records, source pages, reports.
• Logs: parser and provider events.

Right Drawer

Selected record shows claim, section, source URL, snippet, selector/hash where available, and confidence.

Empty, Error, And Blocked States

Research runs must produce either cited records or explicit source/blocked evidence. Docs runs must show crawled sections or why the docs source failed.

Expected Artifacts

Standard artifact contract with research_record.v1 and docs records.

Controls

Run controls, scope filters, section filters, source links, citation context, download, and artifact links.