> ## Documentation Index
> Fetch the complete documentation index at: https://docs.befailproof.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Evaluate agents

> Score every finished session with evaluations you define: hosted Python checks, or LLM judges in your own worker.

An evaluation scores a finished agent session. When a session ends, every enabled evaluation that applies to it runs and records what it found, with reasoning you can read beside the trace:

* a **score** from 0 to 1, optionally marked passed or failed
* a **metric**, such as a count, a duration, or a cost, with its unit
* an **assertion**, which passed or did not

## Two kinds of evaluator

|          | Hosted Python                                        | Your own worker                                                              |
| -------- | ---------------------------------------------------- | ---------------------------------------------------------------------------- |
| Written  | In the dashboard, under **Analyze → eval authoring** | In Python, with the [Evaluator SDK](/reference/evaluator-sdk)                |
| Runs     | On Failproof AI's managed evaluator, in a sandbox    | On your infrastructure                                                       |
| Best for | Deterministic, code-based checks                     | LLM judges, model calls, packages, secrets, network access, heavy processing |

Hosted Python is deliberately small: one expression, no imports, no network. Anything that needs a model — an LLM judge scoring whether an answer was relevant, say — runs in your own worker instead. Neither kind needs an inbound connection: workers claim finished sessions and submit results over outbound HTTPS.

## Each organization evaluates its own agents

Evaluations belong to the organization that defines them. Each organization on an instance writes its own — its own checks, conditions, thresholds, and labels — versions and deploys them without affecting any other, and sees only its own results. Filter those results by agent, environment, evaluation, and time, or ask the assistant about them.

## From first draft to live scores

<Steps>
  <Step title="Write it">
    Describe what to measure and let the assistant draft it, or write it yourself. See [Write an evaluation](/evaluations/write).
  </Step>

  <Step title="Test it">
    Run it against real sessions before it goes live; nothing is stored. See [Test an evaluation](/evaluations/test).
  </Step>

  <Step title="Deploy and version it">
    Deploy an immutable version, publish new ones as it evolves, and roll back to an earlier one. See [Deploy and version](/evaluations/deploy).
  </Step>

  <Step title="Read the results">
    Chart scores over time, compare agents and environments, and ask the assistant. See [Read evaluation results](/sessions/evaluations).
  </Step>
</Steps>

Evaluation runs forward: a version deployed now scores the sessions that finish from now on. To score sessions you already have, [backfill them](/evaluations/deploy#score-sessions-you-already-have).