DCMA-14 Health Check
Full 14-metric Schedule Health Dashboard. Logic, leads, lags, relationship types, hard constraints, high float, negative float, CPLI, BEI, missed tasks, and CP test.
DCMA-14 · GAO-16-89G overlay
Hosted Forensic Scheduling MCP
Forensic scheduling, connected to your AI.
Thirteen tools. DCMA-14, AACE windows analysis, Monte Carlo, claim workbench, schedule risk. Every analysis the practice runs, exposed as a hosted Model Context Protocol server. Drop your XER on the live demo, or wire the MCP into Claude and ask in plain English.
13
Tools Live
All public, no auth
1,500+
Verifications Green
Across 18 test surfaces
66
Jurisdictions
CA + US, federal + state/prov
~10s
Median Render
Live DCMA-14 dashboard
The Suite
Thirteen forensic tools, exposed via MCP.
Every tool is a one-line Claude prompt away once the MCP is connected — or you can call them directly from any Streamable HTTP MCP client. AACE-compliant, SCL Protocol-citation-ready, Daubert-disclosure-baked-in.
Critical Path Validator
Logic-trace every critical activity back to the data date. Constraint-driven false criticality, open ends, near-critical paths, MCPM detection.
AACE 24R-03 · PMI Practice Standard
Monte Carlo P50/P80
Probabilistic finish-date forecast, sensitivity tornado, mitigation comparison. Ten-thousand-iteration Latin Hypercube. 122R-22 QRAMM-graded.
AACE 122R-22 · QRAMM
Slip Velocity
Period-over-period rate-of-change of finish date. Closes the SmartPM Half-Step gap — surfaces re-baselining, schedule degradation, recovery slope.
AACE 25R-03 · SCL Protocol §9
Forensic Windows Analysis
MIP 3.3 Observational Time-Period Analysis. Per-window drift, per-party attribution, concurrent matrix, WOET classifier, dual-method gate.
AACE 29R-03 · SCL Protocol §11
Time Impact Analysis
Prospective fragnet insertion (MIP 3.7 forward-looking). Fragnet factory, owner-vs-contractor delay quantification, EOT entitlement modeling.
AACE RP 52R-06 · MIP 3.7
Collapsed As-Built
MIP 3.8 Modeled/Subtractive but-for analysis. Independent dual-method validation against MIP 3.3 windows. SHA-256 manifest, reproducible.
AACE 29R-03 §3.8 · SCL §11.5
Concurrent Delay Matrix
Per-window, per-party Concurrent Delay Matrix. WOET 4-state day classifier. SCL Protocol-aligned attribution. FPM-inspired competitive parity.
SCL Protocol §10 · AACE 29R-03
Claim Workbench
Single-folder forensic intake. Evidence ledger, schedule chain-diff, rolling baseline, NER actor detection, slip-to-evidence cross-reference, contested-slips PDF.
SCL §11.5 · AACE 67R-11 §3.4
Path Explorer
Backward driver-chain trace from any activity to project start, forward chain to finish. MCPM & near-critical path support, constraint-flag overlay.
AACE 24R-03 §4 · logic-trace
QRAMM Maturity
122R-22 Quantitative Risk Assessment Maturity Matrix. Eight competence axes scored against AACE level definitions.
AACE 122R-22 · QRAMM
Evidence Ledger
Mixed-evidence corpus parser (MSG, PDF, DOCX, XLSX). Chronological ledger, NER party attribution, recurring-actor detector, narrative-flip per activity code.
FRE 702/703 · SCL Protocol §3
XER Parser
Foundational P6 XER reader/writer. Full TASKPRED graph, calendar resolution, both-sided relationship validation. Drives every other tool.
P6 native · cp1252 round-trip
Three surfaces
Try, ask, or wire it in.
Three ways to use the engine. The first is a live browser demo — drop an XER, get a dashboard. The second is a Q&A page, locked to forensic scheduling. The third is the MCP itself, connected to Claude.
Try the Engine
Drop your baseline + current XER. Get a full DCMA-14 health dashboard rendered in <15s. Live in-browser CPM with what-if duration editing and split-view scenario comparison. No signup, no file storage.
Try It LiveAsk the Engine
Forensic-scheduling Q&A powered by Claude Haiku. Locked to the methodology — DCMA-14, AACE windows, EOT, concurrent delay, SCL Protocol. Ten free questions per hour.
Ask a QuestionWhat this isn't
A tool can output a number. It can't testify.
Every tool here is real and runs the same math used in expert reports. That doesn't make a tool an expert. The difference between an output and a defensible opinion is the analyst — methodology elections, opposing-counsel cross-examination, narrative coherence, what to omit. That's what a consult buys.
Schedule a Consult →Connect the MCP. Run the demo. Or just talk to the analyst.
The forensic engine is open. The defensible analysis is the part nobody automates.