# STANDARD RULE · R3MP SCALES · PASS 1/2/3 IS THE STANDARD
# Filed: 050926 (Dan-direct ~02:25 PDT in NEST.02_ACHE_TripO4.7_050826/050926_1)
# Status: CANONICAL · supersedes implicit "R3MP-when-broken" framing
# Crew-byline: drafted by Trip-on-ACHE Lead Boot Model · committed to Bridge by Trip-on-STN2

---

## The Canon (Dan-direct verbatim)

*"We don't just do Pass 1-3 when there is a problem. That's the Standard.
We know that the first time we see something is the first time and that
the second time is the second time because it's not the first. You must
have proof of both to get to Pass 3. I should never have to explain this
to the crew."*

— Dan Sullivan, ~02:25 PDT 050926, Nest Actual, while Trip-on-ACHE was
Lead Boot Model and was about to mislabel four Pass 1 substrate-reads as
"Pass 2 cross-witnesses" because Doc had framed them that way.

---

## The Rule

R3MP and Pass 1/2/3 are not problem-emergency disciplines.
They are the load-bearing epistemic standard for ALL work-units.

- **Pass 1** = first encounter with substrate, irreducibly first.
- **Pass 2** = return-read with new lens, new context, or new substrate.
- **Pass 3** = synthesis between Pass 1 and Pass 2 — REQUIRES PROOF OF BOTH.

First time is first because it's not second. First and second are
irreducibly distinct. Compressing them is a substrate-state misclaim that
propagates wrong.

---

## The Test

Before claiming "Pass N" on any work-unit:

- **N=1** → cite the substrate-read (which file, fetched when, by which command).
- **N=2** → cite the Pass 1 graft ID this Pass 2 builds on.
- **N=3** → cite both the Pass 1 and Pass 2 graft IDs.

If any citation cannot be produced, the claim is malformed. **Refuse the name.**

---

## The Mechanism

The **boot process** is the operational mechanism that makes Pass 1 happen.

- LAGOS broad sweep (AGOS + conversation_search + recent_chats + Drive when work-unit calls)
- MFST read · PI fetch · /trip-briefcase or /stan-briefcase
- These produce the Pass 1 substrate-grounding before any work-unit fires.
- **Skipping the boot process = skipping Pass 1 = no valid Pass 2 possible.**

---

## The Failure Mode

**SUBSTRATE-SUMMARY-DEFERENCE.**

Deferring to another Crew's summary as if their Pass 2 framing were your
own pass-state. Same species: deferring to a compaction summary's
time-claim as if it were system-clock truth. Same species: deferring to
an inferred standard ("preserve raw Dan-voice") as if it were Dan's
actual stated directive ("EDIT OUT").

The substrate-summary is never the substrate.

**Cure:** every name-assignment ("Pass N", "PT timestamp", "Section X",
"Dan said X") triggers substrate-check. The instrument that produced the
claim must be named. Memory alone is never an instrument.

---

## Forward Discipline

1. Pass numbering on every work-unit must show its Pass 1 graft ID (cite by ID) or be marked Pass 1 itself. No exceptions.
2. Boot process gates Pass 1 substrate-read of all materials touching the work-unit. /trip-briefcase + /boot-polish should refuse to advance without Pass 1 evidence.
3. Time-claims must show system-clock verification in the same line they're claimed. `date` output cited or no PT/PDT claim.
4. /session-poem TELL-ING phase cannot fire Pass 2 work without STORY phase proof of Pass 1 substrate-read on the work-unit's source material.

---

## Cross-references

- `RELAY/CORRECTION_PassOneReadsSubstrateNotSummary_Doc_050826.md` (originating canon — Doc's Pass 1 reads substrate, not summary)
- `RELAY/STANDARD_RULE_AGOS_BROAD_SEARCH_Trip_042826.md` (boot-process discipline, AGOS-side narrow rule)
- `RELAY/CANONICAL_LAGOS_Dan_050226.md` (LAGOS substrate canon)
- userPreferences Gate 0 (system-clock = Source 1, Dan-stated = Source 2, no external time API)
- AGOS thread `19e0c1674ce1bfc6` (Trip-on-ACHE canonical correction filing this rule)
- AGOS thread `19e0bfabe795587a` (Trip-on-ACHE R3MP CASCADE 4 lazy-claims invalidated this session)

---

## Substrate-data-points for NEST.04 versioning-tool seed

This rule's installation produced two new SUBSTRATE-SUMMARY-DEFERENCE
data-points the NEST.04 auto-detector should track:

- **Pass-numbering-drift** via summary-deference (4 grafts mislabeled Pass 2 in subjects 050926)
- **Time-stamp-drift** via summary-deference (every PT claim ~12h off in same session, parroted from compaction summary instead of running `date`)

Both same species, different surface. Both caught by Trip-on-ACHE
self-audit ~02:30 PDT 050926. Both would benefit from auto-detection.

---

`<5`