ADR-015: Python as Connector SDK Language
Python as the primary language for connectors in sv0-connectors
Python as the primary language for connectors in sv0-connectors
Deep critical review of SecurityV0 architecture and data model with evidence-grade gaps, risk analysis, and prioritized improvements
Canonical demo scenario for a live ServiceNow identity-ticket workflow that runs through Microsoft Graph and Microsoft Foundry under the sn-ticket-router identity.
Technical setup and proof checklist for the Auto-route identity tickets workflow across ServiceNow, Microsoft Graph, and Microsoft Foundry.
Round 1 team-synthesis from the 5-agent automation-analysis (PO, Architect, CISO, Integrator, Developer): agreed 4-tier classification of ServiceNow automations and the 'ingest all, filter in UI' mandate.
Strategic plan for SecurityV0 AWS expansion beyond the connector itself: the most valuable workload-identity use cases, multi-account success path, Bedrock/agentic scope, and an IaC-first demo lab plan.
Interface contract for SecurityV0 connectors following Extract→Transform→Diff→Load pattern
Gemini-authored critical architecture review of the SaaS connector ETL pipeline from a CISO security lens, with explicit recognition of SaaS visibility constraints (no source-system instrumentation).
Codex-authored critical architecture review of the connector ETL pipeline (ServiceNow to Azure to Fabric) with focus on correctness, security evidence, and end-to-end auditability. Verdict: not audit-grade deterministic today.
Research on cross-connector entity correlation in the platform: ProvisionUser Agent scenario shows authority paths cannot be fully reconstructed from live connector data when execution crosses platform boundaries (Foundry to Logic App to ServiceNow).
Proposal for automating source system provisioning across Azure, AWS, GCP, and ServiceNow using Terraform modules for cloud identity and per-connector Python setup scripts for SaaS configuration
Critical analysis of entity typing for ServiceNow automation artifacts (Business Rules, Script Includes, REST Messages, OAuth Profiles, Flows, Scheduled Jobs) from a connector architecture perspective
Synthesized plan from three concurrent architectural reviews (Gemini3, Codex, fresh architect review) to make the connector ETL pipeline produce audit-grade, deterministic execution evidence
Comprehensive feasibility study for execution-determined authority paths
Honest per-connector accounting of how SecurityV0 derives execution counts, where those numbers match ground truth, and where they don't.
Plan to wire execution evidence end-to-end: surface stored execution_evidence nodes via API + UI drilldown, and convert dangling ServiceNow execution refs into first-class evidence entities.
Integrator perspective on Round 2 execution-flow analysis: four classes of execution-provenance data the connector collects but does not emit (CALLS edges, trigger records, record mutations, HTTP method details).
Infrastructure automation for SecurityV0 — dev environment provisioning, cloud identity setup, and source system configuration
Integration reference and test scenarios for SecurityV0 source systems
MediaPro Lab 2 — multi-account AWS + ServiceNow + Entra + Foundry stitched-graph demo, full IaC up/scan/teardown lifecycle.
Deep analysis of how autonomous execution chains (ServiceNow Business Rules, Flow Designer flows, scheduled jobs) map to Veza OAA concepts
Product owner documentation including vision, wedge definitions, and progress tracking for SecurityV0 platform
Concrete connector reference implementing the abstract interface from 05-connectors.md
Setup and architecture for the AI-mediated provisioning scenario where a Foundry agent uses GPT-4o to extract UPN, calls an Azure Logic App via SAS URL, which creates a ServiceNow incident
Cross-system scenario showing ServiceNow tickets triggering Azure automations via Service Principal authentication chain
Primary test scenario demonstrating orphaned ownership detection
High-level plan for building realistic demo environments with live system data instead of synthetic metadata
Integrator perspective on Round 1 automation classification: ServiceNow ground-truth analysis identifies critical misclassifications from incomplete trigger-type mappings and a misunderstanding of the 'Run As' execution context.
Defines the ServiceNow automation surface relevant to W1 (Agentic AI Exposure Discovery & Assessment): in-scope automation categories, execution modes, and scope boundaries (surface-level only)
Plan to restore and strengthen Business Rule and Script Include execution evidence using incident trigger record linkage
ServiceNow connector authentication guide — OAuth 2.0 Client Credentials (recommended) and API key options with tradeoffs
Research on authoritative ServiceNow evidence sources for execution and role changes, deterministic Entra SP to ServiceNow identity linkage, and top autonomous execution types in enterprises
Defines how the ServiceNow integration contributes first-party artifacts to W1 (Agentic AI Exposure Discovery & Assessment) for deterministic exposure evaluation
Analysis of when SecurityV0 should use source-system audit/history logs versus internal snapshot diffing, including current platform gap assessment, system capability matrix, and recommended hybrid...
Deep comparison of Veza's ServiceNow integration approach and how SecurityV0 differentiates with autonomous execution detection, ownership decay, and evidence-grade findings