# Intel Discipline Advisor — Conversation Flow Example (Industrial) All companies, signals, and figures are **fictional**. Industrial sibling of [`conversation-flow.md`](conversation-flow.md). The SaaS transcript showed inline-input crediting and a combination selection; this one shows the advisor handling a **physical-world signal**, a **paid-tools question answered honestly**, and the routing logic shifting disciplines because the market emits differently. --- **User:** `Intel discipline advisor — a distributor rep mentioned our component competitor just leased a second building near their plant. Means anything?` **Agent:** Heads-up on how this works: three quick questions, then numbered recommendations. Your invocation already answered the first — this is a **suspected competitor move**, and the signal you've seen is a facility lease (that's MASINT flagging once — resource allocation rarely lies, but one signal is still an anecdote). It also arrived via a distributor rep, which makes it HUMINT-carried — worth noting because secondhand facility news needs the primary record. **Question 2 of 3:** what would this competitor plausibly *do* with a second building — more capacity for what they already make, or something new? If you're not sure, say so; the search can carry both hypotheses. **User:** Not sure. Could be capacity, could be the electrification line they keep hinting at in trade interviews. **Question 3 of 3 — constraints:** free sources only, or is paid tooling (trade-data subscriptions, patent analytics) available? And how much time does this get? **User:** We have an ImportYeti-type trade data subscription. Maybe a day, total. **Agent:** Good — that subscription matters here; this is one of the few questions where paid trade data beats an afternoon of free searching. Here's the triage. Your signal is MASINT-flagged (facility) but secondhand. Job one is *confirming the primary record*; job two is *disambiguating capacity vs. new capability* — and those are different signal chains: > 1. **Confirm + read the building (2 hrs, free)** — local permit records and the business > journal for the lease; permits tell you more than square footage: power/water capacity > upgrades and zoning class separate "warehouse" from "production line." *Mapping logic: > MASINT facility chains — land, power, permits precede equipment by 6-36 months.* > 2. **Customs pull on their component imports (2 hrs, your subscription)** — capacity looks > like *more of the same* inputs; a new electrification line looks like *new* input > categories and new supplier geographies. *Mapping logic: MASINT supply-chain chain — input > mix shifts disambiguate exactly your two hypotheses.* > 3. **TECHINT cross-check (2 hrs, free)** — patent classifications and any standards-committee > appearances around electrification; their trade-interview hints are OSINT ambition — a > filing cluster would be commitment. *Mapping logic: announcements are intent until TECHINT/ > FININT corroborate.* > 4. **The off-ramp, honestly:** if the real worry is "will they out-capacity us on existing > products," your sales ops win/loss data on lead times answers it faster than any registry — > ask whether you've been losing on delivery dates first. > > Reply 1, 2, 3, 4, or a combination like "1 and 2". With a day: 1 and 2 fit; add 3 only if 2 > shows new input categories. (Progress: triage complete.) **User:** 1 and 2. If the imports show new categories, I'll come back for 3. **Agent:** Exactly the right sequencing — you've just run the confidence-stacking ladder yourself: MASINT-primary confirmation, then a second independent MASINT chain, holding TECHINT in reserve as the escalation trigger. Two agreeing channels = working hypothesis, enough to assign someone; three = brief leadership. One page to carry it: want the triage card ([`template.md`](../template.md)) filled in with the route, the escalation trigger, and a review date? --- ## Why this example works - **The routing shifted with the physics.** The SaaS transcript routed to SIGINT sweeps and review mining; the same advisor, given a facility signal, routed to permits, customs data, and patents — the mapping table's logic visibly responding to what the market emits, which is the lesson the advisor exists to teach. - **The signal's *carrier* got read, not just its content.** A distributor-relayed facility lease is HUMINT carrying MASINT — so recommendation 1 is *confirm the primary record*, the step secondhand signals always need and usually skip. - **The paid-tool answer was honest in both directions:** the trade-data subscription genuinely changes recommendation 2, and everything else stayed free — no tool-maximalism. - **The user left running the stacking ladder unprompted** ("if imports show new categories, I'll come back for 3") — the by-your-third-session-you-won't-need-me goal, caught on transcript.