Frame the citation-first research assistant experiment

This lab builds citation-first research assistant that maps every claim to evidence IDs. The human in the loop is analyst compiling a brief on bike-share policy from a fixed corpus. Scope is intentionally narrower than “make AI reliable”: you will prove one oracle — brief on fixture corpus: ≥ 0.95 claims cited; fabricated ID rate 0 — and one invariant — claim without evidence → abstain or drop claim; no orphan sentences. Record non-goals in your notes so a later change cannot silently expand authority. The incident mnemonic for the chapter is CAP-RESEARCH-FAKE-ID-3; design as if that ticket is already written and you are filling evidence.

Turn the oracle into a table: input fixture, expected observable, prohibited side effect, owner, latency/cost ceiling. Separate model taste from software correctness — transport, auth, parsing, and termination must be deterministic even when generated text varies. Primary metric family: claim_citation_coverage and fabricated_id_rate. Averages without a denominator or revision label do not gate release. Fake external dependencies in unit tests; live calls wait until fakes pass.

Before implementation, encode a failing check that would have caught model invents kb://minutes-2099 citation. That failure is the pedagogical north star for later pages: contracts reject it, happy path never performs it, validation asserts it, failure-handling contains it, observability detects it, security-ops prevents privilege tricks around it, and mastery replays it in a drill. Endpoint under study: POST /v1/research/brief.

Capacity note for planners: estimate peak demand on POST /v1/research/brief and the cost ceiling for a failed retry storm. Write the abort conditions — unbounded spend, cross-tenant leakage, or inability to roll back — before you enjoy the first green test. Prefer synthetic fixtures shaped like production over anonymized production dumps you cannot share in class. When you are tempted to widen scope, re-read the oracle (brief on fixture corpus: ≥ 0.95 claims cited; fabricated ID rate 0) and cut features that do not serve it. The teaching outcome is judgment under constraints: analyst compiling a brief on bike-share policy from a fixed corpus gets a trustworthy control, not a kitchen-sink framework. Keep the language of release decisions: promote, hold, or roll back — never “see if it gets better.”

Decide what will live in version control on day one: fixtures, contract markdown, and a failing test name. Write the cost ceiling as a hard number with currency and period. If the lab involves clusters, name the non-prod context you will use and forbid prod kubecontexts in scripts. Capture the baseline metric once before changing code so later gains are comparative. Refuse tools that hide the request path behind magic macros until the oracle is green on fakes. Your README section for this page should be five lines or fewer and still falsifiable. 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.