method

the lanes

the corpus was produced by two parallel agent research lanes — history and people, and coverage plus the idea — each writing a lane report against one shared scratch source catalog before synthesis. the pattern is the company-research formula used across this site: lane-prefixed source ids frozen upfront, per-claim confidence, and a declared deduplication owner.

the evidence rules

1. every event, quote, and claim carries at least one source id.
2. dates keep their precision — a year, a month, or a day — and approximate dates are labeled rather than sharpened.
3. confidence stays honest: confirmed, reported, and inferred are distinct claims, not stages of the same claim.
4. sources are typed by role — primary, interview, reporting, archive, community — so a claim's foundation is visible before it is read.
5. sentiment labels describe the coverage, not the truth of it.

known gaps

1. no public funding record was found for moving castles gmbh — the registry shows a single managing person and no investors; how the lab is paid for is unknown.
2. zero's authorship is not fully attributed — the report is unsigned "mc," and the hugging face org membership only partially maps to known names.
3. the current headcount is not verified — the march-2023 de balie billing is the last roster snapshot; the github org shows two public members.
4. zero's eval numbers are self-reported — llm-judged by claude opus 4.8 on the lab's own harness; the hf card's 45.4% versus the report's 45.2% discrepancy is preserved, not reconciled. no third-party replication exists as of release day.
5. gvn's legal name is not public — the pseudonym is respected by the coverage itself and recorded here as a boundary, not a mystery to solve.
6. the original "clearnet paranoia" essay and the mirror tcm interview are dead links; both survive through secondary descriptions and a chinese translation.
7. thiscursedmachine.fun no longer resolves — whether the game is still playable is unverified; our degenerate town remains "under contruction" [sic].
8. the daemons manifesto and roles page are future-dated relative to most of the catalog; late-2026 sources are marked rather than normalized.
9. the x corpus is partial — engagement figures are platform-displayed counts, not audited metrics.

corrections

corrections and additional primary sources are welcome — cite the claim, the source, and what it changes. the catalog is designed to be audited, not just read.