feature · r2-30-evaluator-critic

Evaluator / critic

Taxonomy

M

5 tags · 8 out · 8 in · 3 artifacts · 1 gaps · 1 corpus docs

Catalog tags

group
r2
ev:inferred
maps.dm
absent
status:absent · ev:inferred
maps.technologies
8
ev:primary
slug
r2-30-evaluator-critic
ev:primary
title
Evaluator / critic
ev:inferred

What it is

How should evaluation stay separate from execution — actor proposes, critic scores, human decides on high-stakes paths — without self-approval or unbounded retry?

Why it matters

This topic was frozen in the corporate agentic OS research corpus (wave 10) to answer architecture and measurement questions that sit under the locked loop (observe→understand→predict→intervene→resolve→verify→learn). Stage-1 criticality: M. Research/Spec/Platform readiness: R5/S4/P1.

Related features

Testing, Debugging & Optimization, Independent post-condition checks, Human Oversight & Guardrails

Catalog links

Graph edges for this feature — technologies, category lens, companies, DM.

Technologies

features/r2-30-evaluator-critic/README.md

Artifacts

  • dm_mapabsent
  • readmepresent · features/r2-30-evaluator-critic/README.md
  • technologies_mappresent · features/r2-30-evaluator-critic/technologies.md

Gaps

  • G_DM_CROSSWALKsev 3

    feature missing document-management.md

Out · maps_to (8)

  • abm-agent-behavior-miningtechnology · primary · features/_source/feature-technology-map.json
  • agent-evaluationtechnology · primary · features/_source/feature-technology-map.json
  • deepevaltechnology · primary · features/_source/feature-technology-map.json
  • hal-the-holistic-agent-leaderboardtechnology · primary · features/_source/feature-technology-map.json
  • pm4pytechnology · primary · features/_source/feature-technology-map.json
  • promptfootechnology · primary · features/_source/feature-technology-map.json
  • ragastechnology · primary · features/_source/feature-technology-map.json
  • trulenstechnology · primary · features/_source/feature-technology-map.json

In · maps_to (8)

  • abm-agent-behavior-miningtechnology · primary · features/_source/feature-technology-map.json
  • agent-evaluationtechnology · primary · features/_source/feature-technology-map.json
  • deepevaltechnology · primary · features/_source/feature-technology-map.json
  • hal-the-holistic-agent-leaderboardtechnology · primary · features/_source/feature-technology-map.json
  • pm4pytechnology · primary · features/_source/feature-technology-map.json
  • promptfootechnology · primary · features/_source/feature-technology-map.json
  • ragastechnology · primary · features/_source/feature-technology-map.json
  • trulenstechnology · primary · features/_source/feature-technology-map.json

Corpus tags (document_tags)

feature
r2-30-evaluator-critic
kind
feature_readme
quality
ok
slug
r2-30-evaluator-critic
technology
abm-agent-behavior-mining
agent-evaluation
deepeval
hal-the-holistic-agent-leaderboard
pm4py
promptfoo
ragas
trulens

Corpus documents (1)

feature_readme · 1

  • T30 — Evaluator / Critic Agent Design · ok

    features/r2-30-evaluator-critic/README.md

Research path: features/r2-30-evaluator-critic