# Start Here

> A plain-language introduction for people and AI collaborators.

Truth Machine is a way to maintain shared knowledge that stays explainable as
new evidence arrives, people disagree, and decisions change.

Use it when a group needs more than “the latest note”: people must be able to
see what the current account is, what supports it, and how it was changed.

## The essential idea

- A source claim is **evidence**; it is not automatically the team’s current
  answer.
- Authority is specific. Being confident, senior, or important does not make a
  source authoritative for everything.
- A proposed change says exactly what would change, why, and what evidence
  supports it.
- Review applies to that exact change. Stale, invalid, or partly applicable
  work does not quietly become current.
- The resulting account keeps its reason and ancestry with it.
- Unknowns and disagreements stay visible.
- An AI collaborator can help substantially, but does not gain authority by
  sounding convincing.

## Is this a fit?

The architecture is worth considering when all of these are true:

1. people act on durable shared claims;
2. sources can be incomplete, conflicting, or differently authoritative;
3. changing the current account has real consequences; and
4. the earlier account and reason for change need to remain recoverable.

If that is not your situation, use the simpler tool that fits. Truth Machine
is a way to introduce clarity where the cost of ambiguity is already real.

## When you need precision

The normative specification is under [`/specification/`](/docs/truth-machine/specification/).
Requirements use stable identifiers such as `TM-EVD-001` and the BCP 14 terms
**MUST**, **MUST NOT**, **SHOULD**, and **MAY**. Conformance applies only to
those identified requirements—not to examples, profiles, patterns, or this
orientation.

## Optional practical material

- [Git-native profile](/docs/truth-machine/profiles/git-native/) — one concrete implementation
  using commits, text packets, and an offline reader.
- [AI Peer profile](/docs/truth-machine/profiles/ai-peer/) — how to make a truth system safely
  inhabitable by an AI collaborator.
- [Facets](/docs/truth-machine/patterns/facets/) — disposable, non-authoritative projections at an
  outside boundary.

None is required for core conformance.

## Machine-readable entry points

The published site provides:

- `/llms.txt` — curated reading map;
- `/llms-full.txt` — complete documentation context;
- `/context/core.md` — compact normative context;
- `/manifest.json` — versions, classifications, paths, and content hashes;
- page-level Markdown at each HTML route plus `index.html.md`; and
- `/schemas/` — reference packet schemas.

All are generated from the same documents as the human site. This lets a person
and an AI collaborator start from the same maintained material.