Reference

Licklider Reference is the product-facing technical companion to the task-oriented Docs section.

Use this area when you need to answer questions such as:

• Which analyses and figure families are currently supported?
• Why does the engine make a particular default choice?
• What gets exported, versioned, or recorded in audit metadata?
• Which warnings are advisory, and which ones block claim-bearing export?
• Where does automated checking stop and researcher judgment begin?

Sections

Engine and Support Matrix

Authoritative scope snapshot for supported statistical methods, figure families, and the current boundary between active and legacy implementations.

Implementation Details

Rationale for engine defaults, route and engine responsibilities, and the split between orchestration in JS and numerical execution in Python.

Schemas and Exports

Reference for export surfaces, audit metadata, and how figure results are packaged for downstream reporting and reproducibility.

Versioning and Change Policy

Reference for ruleset and engine versioning, reproducibility guarantees, and how breaking computational changes are communicated.

Limitations and Guard Interpretation

Explains the boundary of automated checking, known blind spots, and how to read warnings, confirmations, and export blocks correctly.

Recommended reading order

1. Start with Engine and Support Matrix to understand current support boundaries.
2. Continue to Implementation Details to understand why the defaults behave the way they do.
3. Use Schemas and Exports when you need machine-readable output details.
4. Review Versioning and Change Policy before relying on long-term reproducibility guarantees.
5. Keep Limitations and Guard Interpretation nearby when you need to interpret warnings conservatively.