# v100.0.22 research integration map

## Evidence and attribution

- Store claim status, confidence, source reliability, and review date separately.
- Use an attribution ladder: content, coordination, identity, sponsorship, intent, reach, effect.
- Count independent origins rather than repeated headlines or reposts.
- Preserve corrections, narrowing updates, and supersession history.

## Movement and organizational claims

- Treat Anonymous as a reusable label and networked identity, not a presumed single organization.
- Distinguish symbols, public supporters, campaign participants, channel administrators, technical cells, and legally attributed defendants.
- Model formal authority and informal influence separately in leaderless networks.

## Oversight, privacy, and remedy

- Compare oversight by powers, access, publication, complaint, and remedy pathways rather than names alone.
- Keep observed metadata, derived inference, and proven identity separate.
- Minimize collection and retention; expose purpose, provenance, correction, appeal, and uncertainty for consequential inferences.

## Site and persona integration

- Starter Packs now include research-backed operating lenses and output checks.
- Persona pages expose voice, reasoning, relationship, boundaries, profile dimensions, examples, and related workflows rather than summary-only cards.
- The research library and API include the new evidence-and-movements group.

## Safety

No operational cyber instructions, illicit-data links, private-person identification, surveillance evasion, or real-world targeting are promoted from the reports into active site or persona memory.