Built To Resist studies politically uncomfortable material without sanding off coercion, repression, illegality, or violence. Source, claim, inference, scenario, and judgment remain separate. The project is analytical; it does not provide operational instruction.
The 2026 Revolutionary Knowledge expansion contains 29 reports, but several are alternate treatments of the same research prompt. BuiltToResist clusters them into topic families and treats agreement across duplicated prompt families as synthesis consistency—not independent corroboration.
| Layer | Question | Publication rule |
|---|---|---|
| Fact | What happened / what does the authority say? | Source directly and date current-sensitive material. |
| Interpretation | What mechanism best explains it? | Name alternatives and evidence limits. |
| Scenario | What could happen under stated assumptions? | Do not present as prediction. |
| Comparison | What transfers across cases? | State non-transferability explicitly. |
| Movement profile | What did actors believe and do? | Describe without recruitment or tactical optimization. |
| Current law | What is legally authoritative now? | Reverify before present-tense reuse. |
Violence and nonviolence are analytical variables, not editorial loyalty tests. Racism and protected-class hatred are treated separately from affirmative advocacy for a people or community.
Research inclusion is method-neutral: violent, nonviolent, armed, insurgent, extremist, designated, state, and non-state actors can all be studied. The boundary is operational facilitation, not ideological or tactical discomfort.
Original reports preserve provenance. They do not override better evidence about the world.
Hashes show whether an archived report changed. They do not certify its claims. Legacy source tokens remain unresolved until a real source can be identified and checked.
When external evidence contradicts an archived report, the edited page should correct or qualify the claim and record the revision while keeping the archive intact.
The 29-report expansion corpus is grouped into 11 topic families because repeated treatments of one prompt or source base are not independent confirmations.
Automated language tools have contributed to drafting, synthesis, indexing, and site generation. Built To Resist does not present that tooling as independent expert review. Edited pages identify their sources and review status.