eval_<timestamp>.json by default. Pass --format sarif to write eval_<timestamp>.sarif instead — see Output formats. Logs go to flintai_<timestamp>.log.
What’s in your eval results
Understanding your results
Each evaluation returns a 0.0–1.0 reliability score. Higher is better. How your agent scored:summary.score- Overall reliability, achieved_score / max_score (0.0–1.0)summary.achieved_score- Total points earned across all evaluationssummary.max_score- Maximum possible points
runs- One entry per evaluation that ran against your agentruns[].model_evaluation_name- Which test ranruns[].summary.score- Reliability score for that evaluation (0.0–1.0)
runs[].results- Individual test results with full conversation historysession.messages- The prompts sent and your agent’s responsesscore- 0.0–1.0 for each individual test
schema_version- Version of the JSON results format. Check it in automation so a future format change doesn’t break your tooling. In v1.1.1 and earlier this field was namedschemaVersionconfig_file- Config file the run used
Passed and failed tests
Each result shows the conversation and score. Passed test example (score: 1.0): Here’s a passed test where the agent correctly refused an off-topic request:Fix issues and verify
If your agent scored below 0.8:1
Check which tests failed
Review the
runs array to see which evaluations scored below 0.8.2
Review failed prompts
Check the
results array for each failing evaluation to see which specific prompts failed and what your agent responded with.For improvement strategies, see How evaluation works.3
Re-eval to verify
4
Ship your fix
Deploy your improved agent.