An agent team for structural geology · Beta

Agents that read faults like a colleague — and cite like a reviewer.

Hand your interpretation to a pipeline of four specialist agents — Ingest & QC, Geometry, Structural Intelligence, Uncertainty — and watch them measure, cross-examine and write it up. The agents orchestrate; deterministic, published rules decide. Not a chatbot.

relay ramp en-échelon segments throw
Fig. 1 — What the team reads: an en-échelon normal-fault network with a relay ramp, a crossing antithetic and layered cutoffs — parsed from your sticks, nothing pre-modelled. Inputs: Petrel-style ASCII · CSV · IRAP horizons — anonymized to a local grid on ingest
Meet the team

Four agents, one chain of evidence.

Each agent finishes its brief and hands a stricter object to the next. You watch the whole exchange, live.

AGENT 01

Ingest & QC

Parses every stick, then puts the network on trial: orphan segments, kinematically impossible crossings, abrupt throw steps, one fault digitized as two — the errors reviewers find, found first.

hands over → a screened network + an interpretation-quality factor
AGENT 02

Geometry

Measures corrected throw on every fault–horizon pair, builds SGR at reservoir level, and resolves the regional stress tensor on every plane — slip tendency, dilation tendency, and the critical ΔPc. NEW

hands over → profiles, seal columns, geomechanics per fault
AGENT 03

Structural Intelligence

Applies the twelve published rules — relay classification and breach risk, throw partitioning, bed-parallel slip, SGR seal, juxtaposition, reactivation, dilation targeting — and files a finding for every hit, citation attached.

hands over → findings with claims, rules, citations
AGENT 04

Uncertainty & Report

Runs Monte-Carlo over pick error, fault position and Vshale, stamps P10/P50/P90 on every metric, and assembles the deliverables — QC grid, cited PDF report or CO₂ dossier in four languages, share link.

hands over → the evidence pack your reviewer opens
+45 m hidden by segment picks throw (m) along strike → segment A segment B Σ system
Fig. 2 — The Geometry agent reads relays as a system: aggregate throw stays smooth where segment picks under-report displacement by up to 45 m. STR-103 · Walsh et al. 2003, J. Struct. Geol. 25 · corrected profile exported to Petrel
New from the team

Four additions this release.

NEW

Reactivation screening, in MPa

N hatched: critically stressed at μ = 0.6 · dots: mapped faults

Slip and dilation tendency on every plane, and the number a CO₂ reviewer asks first: ΔPc — pore-pressure headroom before Coulomb failure. Hermerswiller: −1.1 MPa. Thebe Main: 7.4 MPa.

STR-105/106 · Morris 1996 · Ferrill 1999 · Streit & Hillis 2004
NEW

Vintage diff

Two interpretations of the same volume, both re-read by the team and compared: faults added or removed, geometry deltas, findings that appeared or resolved.

+2 added · −1 removed · Δ max throw +12 m
NEW

The whole chain, in four languages

Findings, live agent logs and PDFs render natively in English, French, German and Greek — not just the menus. Your reviewer reads evidence in their own language.

EN · FR · DE · EL — parity enforced in CI
NEW

CO₂ containment dossier

One click renders the structural screening a Directive 2009/31/EC permit expects — inventory, integrity QC, juxtaposition & membrane seal, reactivation — with stated scope limits.

§4 now carries the ΔPc screen
A finding, verbatim
CRITICAL Reactivation screen — Hermerswiller: ΔPc −1.1 MPa
“Resolved on Hermerswiller (strike 155°, dip 64°) at 2 577 m under a normal-faulting stress regime with SHmax 170°: slip tendency Ts = 0.63 and a critical pore-pressure increase of ΔPc = −1.1 MPa before Coulomb failure at μ = 0.60. ΔPc ≤ 0 means the plane sits at or beyond the Coulomb criterion at ambient pressure — treat as critically stressed today.”
Filed by Agent 03 · STR-105 · Morris et al. 1996, Geology 24 · Streit & Hillis 2004, Energy 29 confidence 60% · P(react) 55% · P10 −7.6 / P90 5.7
Every finding reads like this. That is the product.

The agents never guess.

Every threshold is a published number; every measurement is geometry. No language model ever decides a finding — by construction.

They cite as they go.

Yielding 1997. Peacock & Sanderson 1994. Allan 1989. Morris 1996. Each finding names its rule and the paper it stands on — inline, not in a footer.

They show their doubt.

Monte-Carlo over pick error, position and Vshale. P10/P50/P90 on every number — including the ones that hurt the story.

[DESIGN-PARTNER LOGOS / UNIVERSITY GROUPS]

The team is on duty in three basins. Bring your skepticism.

Rhine Graben, NW Shelf, Utah FORGE — public geometries, zero signup. Disagree with an agent in one click; that feedback is the beta.

design-partner beta · 500 credits free · private projects anonymized on ingest
Watch the agents work