Web4Guru AI Operations
Docs · Specialists · Evaluator

Evaluator

The Evaluator is the gate. It does not produce work; it judges it. Every other specialist's output runs through here against a per-task rubric.

What this specialist produces

  • Pass / fail decision with structured critique on rejection
  • Per-specialist rubric application
  • Hallucination and citation checks
  • Voice / brand match check

How autonomy is handled

Rejects with structured critique; never edits the deliverable. The producing specialist re-does the work.

Most often paired with

The CEO frequently delegates upstream and downstream work to: ceo, every-specialist. See the full roster on the specialists index.

See also