What is it?
A public browser interface and read-only endpoints for a curated subset of knowledge, historical and technical records.
Why was it created?
Make technical context discoverable and explainable, with room to distinguish records, facts and speculation.
Who is it for?
- Students and educators
- Researchers
- Read-only AI knowledge consumers
How to use it
- Open Atlas in the public browser.
- Search for a topic or explore displayed records.
- Read each record’s evidence category and date when provided.
- Use the source context to distinguish fact, hypothesis, future scenario or historical assertion.
Available controls or interactions
Public search and record viewing via approved read-only URLs.
How do you know you succeeded?
You locate relevant published records and understand their scope; one record alone is not proof of its factual correctness.
What should you see?
Matching records and metadata from the published Atlas subset.
Play it through your mind
Ask whether a quoted record is a verified observation, a hypothesis or a future scenario before treating it as advice.
A 2050 technology forecast predicts a major breakthrough. Should you report it as an accomplished fact?
Suggested thought move: Classify the record as a conditional forecast and identify its assumptions.
Reveal the reasoned answer (spoiler)
No. Explain that it describes a possible future, not an observed event.
Illustrative reasoning scenario; no particular record asserted.
For read-only AI clients
GET /atlas/api/search for published data; reason about evidence classification and clearly separate unverified claims.
Public GET/HEAD resources:
AI responses must distinguish proposed actions and local reasoning from actually executed browser actions; no owner credentials or private APIs are provided.
Errors, limits and privacy
If it fails
No-result or unavailable states can occur. Narrow or reformulate the query; avoid interpreting a lack of results as evidence of absence.
Data handling
Only approved public Atlas endpoints should be used; the complete vault and internal project files remain private.
Known limitations
- Index completeness is not guaranteed.
- Published claims may be historical, speculative or incorrect.
- No automated external fact-check certification is implied.