Retrieval and embeddings · Lesson 2 of 16 · 7 min

Inspect source documents

Retrieved evidence is only as useful as its source coverage and validity. Inspect the original document identity and revision before reasoning from a fragment.

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-retrieval-embeddings-002.json. Separate the recorded facts from your assumptions.
Hint & worked example
Retrieved evidence is only as useful as its source coverage and validity. Inspect the original document identity and revision before reasoning from a fragment.
The complete case JSON is read successfully.
```
cat /opt/ai-lab/cases/aiops-retrieval-embeddings-002.json
```
Common mistake to avoid: Assume a high similarity score proves the document is current.
2
Isolate the deciding signal
Select the indexed_revision field from the case JSON. What does it establish—and what does it not establish?
Hint & worked example
Refresh the outdated indexed revision before relying on it for current policy.
A successful JSON query isolates indexed_revision.
```
jq .indexed_revision /opt/ai-lab/cases/aiops-retrieval-embeddings-002.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: refresh-policy-source, change-without-evidence, or ignore-the-signal. Save that one code followed by a newline in /home/learner/lab/aiops-retrieval-embeddings-002-decision.txt.
Hint & worked example
Refresh the outdated indexed revision before relying on it for current policy.
The decision artifact contains the correct code, not just a matching command in history.
```
printf '%s\n' 'refresh-policy-source' > /home/learner/lab/aiops-retrieval-embeddings-002-decision.txt
```
Common mistake to avoid: Assume a high similarity score proves the document is current.

Understand the pattern

Retrieved evidence is only as useful as its source coverage and validity. Inspect the original document identity and revision before reasoning from a fragment.

Interpret the evidence

Refresh the outdated indexed revision before relying on it for current policy.

A tempting mistake

Assume a high similarity score proves the document is current.

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-retrieval-embeddings-002.json

{
  "fixture": "Fictional training data; not a live measurement.",
  "document": "policy-17",
  "source_revision": "v5",
  "indexed_revision": "v3"
}

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

$ jq .indexed_revision /opt/ai-lab/cases/aiops-retrieval-embeddings-002.json

"v3"

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

$ printf '%s\n' 'refresh-policy-source' > /home/learner/lab/aiops-retrieval-embeddings-002-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 “refresh-policy-source” to stop being the justified action. Do not silently replace missing evidence with a guess.

Explain it

What does the evidence in “Inspect source documents” justify?Make the change now and gather the missing evidence afterward.Refresh the outdated indexed revision before relying on it for current policy.Assume a high similarity score proves the document is current.

Apply it elsewhere

A colleague concludes: “Assume a high similarity score proves the document is current.” What should you do?Reject that conclusion; return to the case evidence and its stated limits.Accept it because the training command completed.Treat this fixture as a live measurement.

Keep the momentum.

Complete the lab, then explain and apply the pattern.

Next: Review chunk boundaries 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