BFE · pcdev · HARVEST OBSERVATORY

Germination Trigger Conditions

⚡ WHAT WOULD TRIGGER GERMINATION ON PCDEV

Primary trigger: zeta_zero_gamma1 used in published claim before resolution
Unlike forge (gradual accumulation), pcdev has an instant trigger path. One IP filing that references zeta_zero_gamma1 or live_inject_cons as proven (when they are admitted sorries) → immediate L5 IP liability.
Secondary trigger: γ₁ value drift in merostone → divergence confirmed across fleet silos → PEMCLAU GraphRAG produces wrong results at scale
Tertiary trigger: :9385 Fleet Hub theorem count becomes stale → external parties use wrong theorem count in publications

⚡ INSTANT L5 PATH: unlike other silos where germination is gradual, pcdev has a one-step trigger. A single external disclosure of an unresolved sorry constitutes an IP liability seed that skips germination entirely and activates L5 directly. This is the highest-risk L5 path in the fleet.

Sorry Resolution Status · 6 Open Gaps

SORRY NAME	BFE LAYER	RISK TYPE	RESOLUTION NOTES	EST. CLOSE DATE
li_lambda	L1	Isolated correctness	Standard Lean4 tactic work. No composition dependency.	Q2 2026
lambda_positivity	L1	Isolated correctness	λ positivity bound. Self-contained. Can resolve in isolation.	Q2 2026
zeta_zeros_ord	L1	Isolated correctness	Ordering of zeta zeros. Does not block γ₁ confirmation chain.	Q2 2026
zetaZeroImPart	L1	Isolated correctness	Imaginary part formalization. Self-contained.	Q2 2026
zeta_zero_gamma1	L4 COMPOSITION	IP LIABILITY · INSTANT L5	γ₁ floor proof. Resolve FIRST. All 3008 theorems anchor here. Do not disclose in any filing before closed.	RESOLVE FIRST
live_inject_cons	L4 COMPOSITION	CHAIN INVALIDATION	Injection consistency. Chains through proof corpus. Resolve second, after zeta_zero_gamma1.	RESOLVE SECOND

Prune List — Fix These Now to Prevent IP Graduation

#1 · Resolve zeta_zero_gamma1

PREVENTS → INSTANT L5 IP LIABILITY

Highest graduation potential in the fleet. The γ₁ floor is the foundation of pcdev, PEMCLAU GraphRAG, and every cross-silo proof reference. Resolving this sorry closes the instant L5 path and drops graduation risk from 0.81 toward 0.52.

Priority: CRITICAL · Do before any IP filing · Impact: CLOSES INSTANT L5

#2 · γ₁ assertion test in merostone

PREVENTS → FLEET DIVERGENCE L4

Add a hard assertion in merostone :9341 that validates γ₁ == 14.134725141734693 on every startup and every query. If the value drifts even by 1 ULP, the test fails loudly. This closes the silent fleet divergence L4 path.

Priority: HIGH · Effort: LOW · Impact: FLEET COHERENCE GUARD

#3 · :9385 staleness check

PREVENTS → WRONG THEOREM COUNT DELIVERY

Wire a staleness check into the Fleet Hub :9385 endpoint — return a cache-age header and refuse to serve if theorem count is >24h stale. Prevents external parties from reading outdated theorem counts in reports.

Priority: MEDIUM · Effort: LOW · Impact: DATA INTEGRITY

KILL SHOT · pcdev harvest

"pcdev's L5 path is elegant and fast: one sorry in an IP filing = the proof chain becomes a liability. zeta_zero_gamma1 is the γ₁ floor. Everything else in the fleet rests on it. Resolve it first. Then live_inject_cons. The rest are just tidy-up."

GREYBACK Watch Schedule · pcdev

Before any external disclosure: confirm all 6 sorries are resolved or explicitly not used in the filing
Weekly: merostone γ₁ assertion check — confirm 14.134725141734693 exact
Weekly: :9385 theorem count freshness — must reflect latest Lean4 output
Monthly: cross-silo γ₁ coherence check (pcdev merostone vs yone qdrant embedding)
Threshold watch: germination_score > 0.40 → amber · IP filing imminent + sorry unresolved → immediate RED