Instrument evidence trails and abstentions

Instrument citation-first research assistant that maps every claim to evidence IDs so analyst compiling a brief on bike-share policy from a fixed corpus can answer: demand, errors, latency/age, saturation. Emit fields needed by claim_citation_coverage and fabricated_id_rate with bounded labels. Sample successful high-volume traces; keep errors and rollout transitions denser within policy.

Define at least one alert that would fire for CAP-RESEARCH-FAKE-ID-3, with a for/pending window that survives deploy blips. Missing scrape or missing revision labels must not look like health. Include a trace/log example id format you will actually search.

Add a unit/integration check that metrics increment on the happy path and on the model invents kb://minutes-2099 citation branch. Store machine-readable output in CI artifacts when practical.

Cardinality discipline: tenant and revision are usually enough; raw question text is not a label. Exemplars or trace links beat screenshots alone when debugging CAP-RESEARCH-FAKE-ID-3. Define who owns alert fatigue review. If you export to a vendor, record retention and access. Synthetic probes should use non-sensitive fixtures and still exercise POST /v1/research/brief. Practice the query you will type at 2am once, while calm.

Document the exact PromQL or log query in the runbook stub for this service. Verify histograms have buckets around your SLO target. Add a canary synthetic check that exercises the oracle path every few minutes in staging. Confirm that PII redaction happens before export. Track build/version as a label on the golden signals. Delete noisy debug logs before they become accidental product dependencies. In this chapter the product is citation-first research assistant that maps every claim to evidence IDs, the human stakeholder is analyst compiling a brief on bike-share policy from a fixed corpus, and the incident id you design against is CAP-RESEARCH-FAKE-ID-3. Re-state the oracle in your notes — brief on fixture corpus: ≥ 0.95 claims cited; fabricated ID rate 0 — and keep the invariant visible: claim without evidence → abstain or drop claim; no orphan sentences. Track claim_citation_coverage and fabricated_id_rate as the scoreboard. Surface under change control: POST /v1/research/brief. If you only have forty minutes, finish the fixture for model invents kb://minutes-2099 citation before polishing UI. Promotion language stays ternary: promote, hold, or roll back based on evidence, not hope.

Set a timer for twelve minutes and attempt only the observability concerns for Capstone: research bot with citations. Speak aloud what analyst compiling a brief on bike-share policy from a fixed corpus would see on success versus on model invents kb://minutes-2099 citation. Write three bullet artifacts you must keep: a fixture name tied to CAP-RESEARCH-FAKE-ID-3, a metric query for claim_citation_coverage and fabricated_id_rate, and the rollback or refuse command for POST /v1/research/brief. Then extend the notes with one sentence on how claim without evidence → abstain or drop claim; no orphan sentences would be violated if you skipped this page. Close by restating the oracle in your own words without looking: brief on fixture corpus: ≥ 0.95 claims cited; fabricated ID rate 0.