Evaluations and datasets · Lesson 17 of 18 · 9 min

Archive evaluation evidence

Archived evaluation evidence must preserve the reproducible inputs and results. Sensitive raw examples may need controlled storage rather than inclusion in a public report.

Open the lab Course map

Mastery 0/3Run itExplain itApply it

learner@clairos:~ $

↑ history · Tab complete · Ctrl+C cancel · Esc, Tab to leaveExit 0

Your objectives

Hints are always available. Wrong turns never cost XP.

1
Read the case evidence
Read /opt/ai-lab/cases/aiops-evals-datasets-017.json. Separate the recorded facts from your assumptions.
Hint & worked example
Archived evaluation evidence must preserve the reproducible inputs and results. Sensitive raw examples may need controlled storage rather than inclusion in a public report.
The complete case JSON is read successfully.
```
cat /opt/ai-lab/cases/aiops-evals-datasets-017.json
```
Common mistake to avoid: Keep only the aggregate score and delete the run manifest.
2
Isolate the deciding signal
Select the manifest_saved field from the case JSON. What does it establish—and what does it not establish?
Hint & worked example
Archive the versioned manifest and results without publishing sensitive raw cases.
A successful JSON query isolates manifest_saved.
```
jq .manifest_saved /opt/ai-lab/cases/aiops-evals-datasets-017.json
```
Common mistake to avoid: Using a different case or an unrelated signal as the justification.
3
Record your operational decision
Choose the justified code: archive-bounded-evidence, change-without-evidence, or ignore-the-signal. Save that one code followed by a newline in /home/learner/lab/aiops-evals-datasets-017-decision.txt.
Hint & worked example
Archive the versioned manifest and results without publishing sensitive raw cases.
The decision artifact contains the correct code, not just a matching command in history.
```
printf '%s\n' 'archive-bounded-evidence' > /home/learner/lab/aiops-evals-datasets-017-decision.txt
```
Common mistake to avoid: Keep only the aggregate score and delete the run manifest.

Understand the pattern

Archived evaluation evidence must preserve the reproducible inputs and results. Sensitive raw examples may need controlled storage rather than inclusion in a public report.

Interpret the evidence

Archive the versioned manifest and results without publishing sensitive raw cases.

A tempting mistake

Keep only the aggregate score and delete the run manifest.

All names, measurements, prices, versions, and thresholds in this exercise are fictional training data. The evidence supports this bounded scenario, not a claim about a live provider.

Worked example and expected result

$ cat /opt/ai-lab/cases/aiops-evals-datasets-017.json

{
  "fixture": "Fictional training data; not a live measurement.",
  "run_id": "eval-23",
  "manifest_saved": true,
  "results_saved": true,
  "raw_sensitive_cases_public": false
}

Exit status: 0. The command completed successfully in the training fixture.

$ jq .manifest_saved /opt/ai-lab/cases/aiops-evals-datasets-017.json

true

Exit status: 0. The command completed successfully in the training fixture.

$ printf '%s\n' 'archive-bounded-evidence' > /home/learner/lab/aiops-evals-datasets-017-decision.txt

(no output; exit 0)

Exit status: 0. The command completed successfully in the training fixture.

Try a variation

Name the assumption that would have to change for “archive-bounded-evidence” to stop being the justified action. Do not silently replace missing evidence with a guess.

Explain it

What does the evidence in “Archive evaluation evidence” justify?Archive the versioned manifest and results without publishing sensitive raw cases.Keep only the aggregate score and delete the run manifest.Make the change now and gather the missing evidence afterward.

Apply it elsewhere

A colleague concludes: “Keep only the aggregate score and delete the run manifest.” What should you do?Accept it because the training command completed.Treat this fixture as a live measurement.Reject that conclusion; return to the case evidence and its stated limits.

Keep the momentum.

Complete the lab, then explain and apply the pattern.

Next: Evaluations checkpoint Try an incident drill

Anonymous lesson and mastery progress lasts for this session. Sign in to keep synced lesson XP between sessions; mastery stays device-local. No speed bonus, lost lives, or hint penalties.

Keep learning

Case methodology: evidence, limits, decision