§19 · The moat
What's actually hard to copy
A feed aggregator can be cloned in a weekend. A trusted field-intel commons cannot. The value lives in relationships, contributors, verification history, and institutional trust — and each dimension below is anchored to the §18 build order.
Community reputation
The accumulated standing of the contributor base — who shows up, who is trusted, who gets cited.
Why hard to copy: Reputation is earned over years of public behavior. A new entrant starts at zero even with identical software.
- contributor reputation
- contributor profiles
Field note corpus
The growing body of structured field notes attached to signals, places, entities, and claims.
Why hard to copy: The corpus is path-dependent; you cannot retroactively annotate events that already happened without the contributors who were there.
- field_notes
- field note submission
- note attachment to signal/place
- note status
Provenance graph
Every claim traceable to sources, observations, evidence, and the edits that shaped it.
Why hard to copy: Provenance must be recorded as the work happens. Reconstructing it after the fact is impossible — the audit trail either exists or it does not.
- provenance_events
- transparent change history
- evidence_items
Verification workflows
The corroboration, contradiction, and promotion machinery that turns raw notes into claims.
Why hard to copy: Workflows encode hard-won operational judgment — what to escalate, what to quarantine, what to promote. Copying the UI doesn't copy the operating model.
- verification queue
- corroboration / contradiction workflow
- promoted notes
- basic moderation
Trust history
A durable ledger of who was right, who was wrong, and how reliability has changed over time.
Why hard to copy: Trust history is cumulative and adversarially tested. A fresh deployment has no track record to point at when stakes rise.
- trust_events
- source reliability cards
- simple confidence labels
Maintainer network
The people who keep adapters healthy, moderate edge cases, and absorb operational load.
Why hard to copy: Maintainers are recruited, trained, and retained. The network compounds as members vouch for newcomers — it cannot be spun up overnight.
- maintainer program
- moderation_actions
Partner integrations
Bilateral arrangements with feed providers, responders, and downstream consumers.
Why hard to copy: Each partnership is a negotiation, a contract, and a relationship. They accrete slowly and travel with the institution, not the codebase.
- partner workspaces
- public API
Briefings from living public intelligence
Briefings generated from the corroborated, source-linked commons — not from a static dataset.
Why hard to copy: Useful briefings depend on a fresh, contradiction-aware substrate. Without the corpus and verification layer beneath them, briefings degrade to summaries of strangers' opinions.
- public briefings
- GraphRAG summaries
- claim pages
Contributor identity without forced exposure
Pseudonymous identity with real reputation — contributors can build standing without becoming targets.
Why hard to copy: Designing identity that is durable, accountable, and protective at once requires careful schema, policy, and tooling decisions made early. Bolting it on later breaks existing reputations.
- contributor reputation
- contributor profiles
The public commons itself
The shared artifact — open data, open methodology, open license — that the whole network defends.
Why hard to copy: A commons is a social contract, not a feature. Forks can copy the bytes; they cannot copy the constituency that treats this commons as theirs to protect.
- open license / contributor terms
- governance board
- dataset exports