Agentic AI › Assess
Evaluations
Whether the agent was any good — graded by a person, by another agent, or by what happened to what it produced. A proposal accepted and still standing thirty days later is the strongest signal this table can carry.
Where to find it
- Module: Agentic AI
- Group: Running
- Section: Assess
- Screen: Evaluations
Address in the console: /agentic/evaluations
List · 11 columns · 10 fields on its form · 2 of them required
What the list shows
These are the columns of the list, in the order they are drawn. Any of them can be sorted on, filtered on, hidden, or exported.
| Column | Kind | Field name |
|---|---|---|
| When | Date and time | evaluated_at |
| Agent | Text | agent_name |
| Run | Code | run_no |
| Graded by | Status | evaluator |
| Person | Text | evaluator_name |
| Criterion | Text | criterion |
| Score | Number | score |
| Verdict | Status | verdict |
| Outcome | Status | outcome_kind |
| Value | Amount | outcome_value |
| Comment | Text | comment |
The form
What is asked for when a record is created here. Required fields are marked; a choice drawn from a configuration list is one an administrator can extend without a release.
| Field | Type | Choices | Notes |
|---|---|---|---|
| Run | Choice | Looked up from existing records | Optional: the run being graded, from this agent's runs. Empty grades the agent in general. |
| Agent required | Choice | Looked up from existing records | The agent whose work is graded. |
| Graded by required | Choice | Configuration list | Who judged it. A person: you or a colleague. Another agent: an agent acting as grader (choose it below). What happened next: the result of its work, e.g. a proposal accepted and still standing. A rule: an automatic check. |
| Graded by agent | Choice | Looked up from existing records | The agent that did the grading — usually a reviewer agent set up to check other agents' work. |
| Criterion | Text | What the grade is for, e.g. overall, accuracy, tone. Leave blank for overall. The date is recorded automatically when the evaluation is saved (When). | |
| Score | Number | Out of 100: 0 is the worst, 100 the best, e.g. 80. Optional — Verdict alone is enough. | |
| Verdict | Choice | good, acceptable, poor, wrong, harmful | good: right and useful; acceptable: usable with small fixes; poor: needed real rework; wrong: incorrect; harmful: would have caused damage if acted on. Kept beside Score; neither is worked out from the other. |
| Comment | Long text | Why, in a sentence — the most useful part for whoever improves the agent. | |
| Outcome | Text | For What happened next: what became of the work, e.g. accepted or rejected (filled in by Accept it and Reject it). | |
| Value | Number | What that outcome was worth, in the company currency, e.g. the cost saved. Optional. |
The rules that apply here
Everything on this screen obeys the platform rules rather than rules of its own. The ones worth knowing before you use it:
- Narrowing a list — search, conditions, sort, totals, export and saved views.
- What may be changed — the five grounds a record may not be edited on.
- Companies and business units — what the switcher above this list is doing to it.
- Creating a record — required fields, where choices come from, and how the number is allocated.
- Importing a spreadsheet — this screen accepts one, and it runs through the same rules as typing.