method

the lanes

the corpus was produced by four parallel agent research lanes — company history, the people and their writing, coverage and sentiment, and the file-over-app idea — each writing a lane report against one shared scratch source catalog before synthesis. the pattern is the company-research formula proven on roam-research: lane-prefixed source ids frozen upfront, per-claim confidence, and a declared deduplication owner — the fixes the roam run said the formula needed.

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. the “first line of code, january 31, 2020” claim traces to one fan wiki — plausible, weak, labeled reported.
2. no evidence steph ango worked at ink & switch — an earlier draft said otherwise and was corrected during synthesis. the relationship is intellectual adjacency.
3. financing history can be asserted but not proven negative — no venture, angel, or grant disclosure was found, and none may exist.
4. user and revenue counts are self-reported; obsidian runs no analytics, so even the company measures by proxy.
5. the june 2026 sync price-hike coverage leans on remio.ai, whose sharpest statistics read ai-embellished — the backlash is real, its exact numbers unverified.
6. shida li and erica xu give few interviews; the founder corpus is thin and the asymmetry is recorded, not filled.
7. discord history from the beta era is not publicly archived; early community claims lean on forum threads and contemporaneous posts.
8. the x/twitter corpus is partial — tweet-level material came through archives and interviews, not native search.
9. chinese-community interviews exist in translation on the zh forum; the english-language corpus undercounts them.
10. “$350m valuation” and “$25m arr” claims in 2025–26 secondary coverage are unsourced — treated as reported-speculative and kept out of the record.

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.