ConsultingWorkbench-backed AI security engagements — map, attack, defend, and prove your AI systems.
Scope a Review
Publication DSL

A structured language for reports, deliverables, field guides, and evidence.

The DSL keeps Markdown readable while giving high-value sections real structure: findings, risks, controls, evidence packs, permission matrices, trust boundaries, charts, print notes, and executive decisions.

example

Mini AI Security Deliverable Example

A compact example showing nested Publication DSL blocks for a client-facing AI security deliverable.

example

Mini AI Security Report Example

A compact report example showing findings, charts, evidence, and recommendations.

example

Publication DSL Block Gallery

A single example file demonstrating every official v1 block.

deliverable

AI Product Security Assessment Sample

A fuller sample showing how the Publication DSL can structure a serious AI Product Security review.

deliverable

Enterprise AI Security Evidence Pack Sample

A DSL sample for a buyer-ready enterprise AI security evidence pack.

field-guide

Field Guide Chapter DSL Sample

A sample field guide chapter using the Publication DSL for teaching and advisory content.

Official v1 blocks
:::callout
Callout

A short emphasized note, warning, evidence note, or interpretive aside.

:::stat-grid
Stat Grid

A compact grid of metrics, proof points, or review snapshot values.

:::pull-quote
Pull Quote

A visually prominent quote or thesis line.

:::checklist
Checklist

A checklist of controls, readiness items, or operational tasks.

:::section-opener
Section Opener

A strong section opener with optional kicker and narrative lead.

:::finding-card
Finding Card

A structured security, market, governance, or maturity finding.

:::finding-grid
Finding Grid

A container for multiple finding-card blocks.

:::risk-register
Risk Register

A data-backed register of AI risks, owners, status, severity, evidence, and remediation.

:::evidence-pack
Evidence Pack

A data-backed summary of controls, evidence, implementation status, and ownership.

:::control-map
Control Map

A data-backed control map for frameworks, buyer questions, ownership, and evidence status.

:::permission-matrix
Permission Matrix

A data-backed matrix of agent/tool permissions, action classes, approvals, logging, and owners.

:::trust-boundary-map
Trust Boundary Map

A data-backed trust boundary visualization for users, AI systems, providers, tools, and logs.

:::decision-box
Decision Box

A high-signal executive decision, condition, or recommendation.

:::chart
Chart

A data-backed chart wrapper using the existing chart system where possible.

:::table
Table

A markdown table wrapper with title, print behavior, and lintable identity.

:::artifact
Artifact

A reference to a deliverable, figure, source file, route, or generated asset.

:::page-break
Page Break

A print/PDF page break hint.

:::print-note
Print Note

A note specifically intended to clarify print/PDF behavior or static fallback.