Direct answer
A personal research ecology should strengthen the user’s thinking by making sources clearer, claims easier to inspect, and synthesis easier to revise. It should not hide uncertainty or convert first drafts into unsupported authority.
Component map
| Niche | Specialist | Evidence |
|---|---|---|
| Source intake | Parser and chunker | Complete file coverage |
| Claim extraction | Narrow extraction model | Claim/source alignment |
| Synthesis | Writing model | Coherence and usefulness |
| Critique | Independent reviewer | Missing evidence and counterpoints |
| Citation checker | Deterministic plus model-assisted | No unsupported source references |
| Memory updater | File-backed ledger | Accurate source-to-memory promotion |
Workflow
PROCEDURE research_session(question, sources)
chunks <- INGEST_SOURCES(sources)
claims <- EXTRACT_CLAIMS(chunks)
synthesis <- DRAFT_SYNTHESIS(question, claims)
critique <- CRITIQUE_SYNTHESIS(synthesis, claims)
citations <- CHECK_CITATIONS(synthesis, chunks)
RETURN REVIEW_PACKET(synthesis, critique, citations, unresolved_questions)
END PROCEDUREHuman-strengthening rule
The assistant should make the user better at research: show evidence, preserve uncertainty, invite revision, and keep source notes durable. A good research ecology increases agency instead of creating dependency.
Source reports used for this guide
These reports are preserved verbatim in the site archive. The guide above is an editorial synthesis and may narrow, qualify, or reorganize claims from the source material.