How it works

Foe: Mirage Djinn, an AI answer that sounds right but isn't.

Mirage Djinn

1. Mirage Djinn

   The Mirage Djinn blocks the way.

   What really happensIt stands for an AI answer that sounds right but isn't.

   Summoner: Context 100/100.

2. Mirage

   The djinn conjures a mirage, and the hero swings at nothing.

   What really happensA model can sound sure and still be wrong, so AI features need evals and checks.

   Summoner takes 8. Summoner: Context 92/100.

3. MLflow Skills

   The MLflow kit opens its tracing scrolls.

   What really happensinstrumenting-with-mlflow-tracing adds MLflow Tracing to your Python or TypeScript app and its LLM library.

   Mirage Djinn takes 28.

4. MLflow Skills

   Footprints show where the agent took a wrong turn.

   What really happensanalyze-mlflow-trace and debug-agent read spans and assessments and connect them to your code.

   Mirage Djinn takes 23.

5. Mirage

   The Mirage Djinn uses Mirage.

   Summoner takes 8. Summoner: Context 84/100.

6. MLflow Skills

   The hero sets up judges for the next trial.

   What really happensbuild-a-scorer turns your quality rules into code checks, built-in scorers or LLM judges.

   Mirage Djinn takes 35.

7. MLflow Skills

   The agent is tested before it is trusted.

   What really happensagent-evaluation builds a dataset, runs the evaluation and reports results so fixes can be measured.

   Mirage Djinn takes 14. Summoner: Ward. Mirage Djinn falls.

8. Victory

   The Mirage Djinn falls.