---
type: source-summary
title: 'Source: April 2026 refresh batch (4 updated notebook sources)'
modified: 2026-04-11
tags:
  - april-refresh
  - competitive-update
  - canonical-metrics
  - leadership-brief
  - source
---

# Source: April 2026 refresh batch (4 updated notebook sources)

## What this source is

Four notebook sources added in April 2026 that **supersede or extend earlier versions** of the same content. These are the freshest content in the entire corpus.

| # | Title | Chars | Key topic |
|---|---|---|---|
| 1 | Leadership Brief — April 2026 (Answers Andy's 3 Questions) | 2,320 | Standalone vs integrated, deployment is Squire's decision |
| 2 | Competitive Landscape Update — April 2026 (Supersedes March 2026) | 2,557 | Competitors HAVE shipped AI; Celigo = CRITICAL THREAT |
| 3 | 26-CANONICAL-METRICS-AND-WORDING.md (April 2026 refresh) | 3,675 | **409 suites, 9,364 tests** — newest canonical numbers |
| 4 | 04-TECHNICAL-PROOF.md (April 2026 refresh) | 5,642 | DLP now says 14 patterns with field-gating; test numbers updated |

## Key claims

### 1. Leadership Brief — Andy's 3 Questions (NEW, high adoption-case value)

**Andy** is a Squire executive who asked three specific questions. The brief answers them directly:

**Q1: "Is it a standalone product?"**
A: *"It can be."* SuiteCentral 2.0 runs independently — own dashboard, AI engine, 16-module library, connector framework. Connects to both ERPs. Does NOT depend on SuiteCentral 1.0.

**Q2: "Can Squire incorporate it into existing SuiteCentral products?"**
A: *"Also yes."* Designed as the AI upgrade path for PaymentCentral, VendorCentral, SyncCentral, etc. — the modules Squire already sells to 40+ clients. Adds AI field mapping, governance, reasoning traces, dual-ERP. Architecture supports BOTH paths (standalone or integrated). **Deployment strategy is Squire's decision.**

**Q3: "What is the end goal?"**
A: *"Keep Squire competitive and ahead of the market."* Lists competitors who've shipped AI: Oracle (free AI in 2026.1), Celigo (MCP + AI Assist + agentic), Workato (autonomous agents). SC 2.0 gives Squire a platform it OWNS with governance capabilities none of them have.

→ **NEW wiki page needed**: Andy's 3 Questions is a concise adoption-case framing that directly addresses the "what do we actually DO with this" question. Should be linked from [[pages/entities/squire]] and the executive reading guide.

### 2. Competitive Landscape Update — April 2026 (SUPERSEDES March)

**"This update supersedes all previous competitive landscape documents."** Key changes:

- **Celigo — UPGRADED TO CRITICAL THREAT**: shipped MCP Server, AI Assist, Mapper 2.0 with LLM output processing, 20+ AI-powered templates, agentic automation. Still #1 G2, $79.7M funded, 4,000+ customers. **STILL MISSING**: reasoning traces, confidence scoring, approve-to-apply, DLP/PII, dual-ERP governance.
- **Oracle NetSuite — CONFIRMED CRITICAL**: 2026.1 maturing AI, free MCP tools, first-party metadata access advantage. **STILL MISSING**: reasoning traces, confidence scoring, approve-to-apply, dual-ERP.
- **Workato — NOW RELEVANT** (elevated from "not in register"): agentic AI orchestration, 75,000+ agents, MCP + Google A2A. **STILL MISSING**: ERP-specific governance layer.
- **OpenAI Frontier — HIGH THREAT**: $1M+/yr minimum, 6-month implementation. Squire's mid-market ($5M-$200M) won't adopt.
- **Strategic pivot**: *"Previous claim: 'SuiteCentral is the first to add AI to integration.' Corrected position: 'Everyone has shipped AI integration. Nobody has shipped AI integration governance.'"*

→ **Updates [[pages/concepts/competitive-landscape]]** with the upgraded threat levels and the explicit "first mover → governance depth" pivot.

### 3. Canonical Metrics April 2026 Refresh — NEWEST canonical numbers

The authoritative wording has been refreshed:

- **409 suites** (up from 406 in March)
- **9,364 tests passed across 9,394 total** (up from 9,364/9,394)
- **Full breakdown: 9,198 unit (23 skipped) + 170 integration (7 skipped) + 20 E2E portal**
- Coverage, LOC, module, connector, provider wording: unchanged from March
- "6 production connectors" still includes Oracle (despite CONNECTOR_STATUS saying Beta)
- Claude still listed as "3.5 Sonnet" in the canonical wording (not updated to 4.5 yet)

→ **Updates [[pages/concepts/production-proof]]** vintage table and [[pages/concepts/canonical-metrics]] with the new canonical numbers.

### 4. Technical Proof April 2026 Refresh — DLP 14 patterns confirmed

The DLP/PII section has been updated to match the PII Detection Guide:

> **"14 patterns (SSN, CC, email, phone, intl phone, MRN, API key, JWT, IP, bank account, DOB, passport, driver's license, name — 6 field-gated) auto-redaction"**

This is the first time the canonical technical proof document explicitly lists all 14 patterns AND notes the 6 field-gated ones. Previous versions said "9 patterns." The user (or someone) has already updated this doc to match the codebase.

Also: test numbers have since been rebaselined to the April 19, 2026 canonical — **9,476 passing + 34 skipped = 9,510 total in 419 suites** (per [[sources/26-canonical-metrics-and-wording]]). The earlier "9,283 + 37" snapshot in this batch reflected an intermediate rebaseline point.

## Pages updated by this ingest

**Created** (0 new pages — Andy's 3 Questions content folded into existing pages)

**Updated** (4 existing pages):
- [[pages/concepts/competitive-landscape]] — Celigo CRITICAL THREAT, Workato elevated, strategic pivot to "governance depth"
- [[pages/concepts/production-proof]] — newest canonical numbers (409 suites, 9,364 tests), vintage table updated
- [[pages/concepts/canonical-metrics]] — refreshed to April 2026 canonical (409/9,364/9,394)
- [[pages/entities/squire]] — Andy's 3 Questions: standalone OR integrated, Squire's choice

## Cross-references / contradictions

- **Multiple test vintages in April refreshes**: Leadership Brief says 384/9,099/9,122; Tech Proof refresh says 406/9,283/9,320; Canonical refresh says 409/9,364/9,394. These are different snapshots taken at different times during a period of rapid test growth. The canonical (409/9,364) is the authoritative number per the style guide's governance rule.
- **"First mover → governance depth"**: the competitive update EXPLICITLY corrects the prior "first to add AI" framing. This is a strategic-narrative pivot, not just a wording change. All wiki pages that referenced "first mover" positioning should be checked — though the Brain1 wiki already uses the governance-depth framing per the hook video and talking-points sources.
- **Claude still "3.5 Sonnet" in canonical wording**: the PRODUCTION-VS-DEMO-GUIDE.md shows `claude-sonnet-4-5-20250929` (4.5 Sonnet) but the canonical metrics hasn't been updated. Flag: the canonical wording file is behind the production config.
- **Oracle still "production" in canonical**: CONNECTOR_STATUS says Beta; canonical still says "6 production connectors" including Oracle. Both are in the repo. The discrepancy persists.