Use Cases

Real workflows. Real approval gates.

Four built-in cross-domain workflows and a 6-agent AI council — all running through the same human-approved mutation path. These are not hypothetical. They ship in the community edition.

Incident response

Trigger: New incident created or severity escalation event

When an incident is logged, the workflow activates across multiple domains to build a complete response picture without manual coordination.

01

Classify and enrich

AI reads the incident details, queries ITSM for affected assets, and proposes initial severity and category classifications.

Incidents + ITSM

02

Timeline and evidence

Proposes timeline entries from available data, links related evidence records, and identifies gaps in the evidence chain.

Incidents + Evidence

03

Control gap analysis

Maps the incident to relevant controls, identifies which controls failed or were absent, and proposes control remediation actions.

Controls

04

Lessons learned

Drafts lessons learned entries, links them to the incident, and proposes updates to related risk scenarios.

Incidents + Risks

Approval gates: 4 approval gates — one per step. Each proposed mutation pauses the workflow until a human approves or rejects.

Periodic risk review

Trigger: Scheduled cron job (e.g. weekly or monthly)

A scheduled run walks the entire risk register, evaluates current state against tolerance statements, and proposes updates where the data warrants it.

01

KRI evaluation

Reads all Key Risk Indicators, checks current values against thresholds, and identifies breaches or trending indicators.

Risks

02

Scenario reassessment

For risks with KRI breaches, proposes updated likelihood and impact scores based on current evidence and incident history.

Risks + Incidents

03

Tolerance analysis

Evaluates updated risk scores against the Risk Tolerance Statement. Flags risks that have moved outside acceptable bounds.

Risks

04

Treatment proposals

For out-of-tolerance risks, proposes new treatment actions or escalates existing treatment plans with revised timelines.

Risks

Approval gates: Each proposed score change, tolerance breach flag, and treatment action requires human approval before it modifies the register.

Control assurance

Trigger: Assessment cycle start or event trigger

The AI council reviews control effectiveness across the framework, identifies weak points, and drives evidence collection to close gaps.

01

Effectiveness review

The 6-agent council examines control test results, assessment scores, and linked incident data to produce an effectiveness opinion.

Controls + Council

02

Evidence gap detection

Identifies controls that lack current evidence, have expired artifacts, or have never been tested in the current assessment period.

Controls + Evidence

03

Evidence requests

Proposes evidence collection requests for identified gaps, assigns to relevant owners based on control ownership data.

Evidence

04

Remediation planning

For controls rated as ineffective, proposes specific remediation actions with suggested timelines and links to related nonconformities.

Controls + Audits

Approval gates: Council opinions are presented for review. Every proposed evidence request, remediation action, and status change requires approval.

Policy compliance cycle

Trigger: Policy review date reached or regulatory change event

When a policy is due for review or a regulatory change triggers reassessment, the workflow coordinates updates across the policy lifecycle.

01

Policy review

Reads the current policy version, compares against linked framework requirements, and identifies sections that may need updates.

Policies

02

Control mapping validation

Checks that policy statements still map correctly to implemented controls. Flags orphaned mappings or missing coverage.

Policies + Controls

03

Risk alignment

Validates that policy scope aligns with the current risk register. Identifies risks that lack policy coverage.

Policies + Risks

04

Change request

If updates are needed, proposes a formal change request with draft modifications, submits for the approval workflow.

Policies

Approval gates: Policy changes go through the built-in approval workflow. No policy is published without explicit human sign-off.

Real queries, real responses

What actually happens when you ask a question.

Query

Show me the top risks

Returns 15 risks with scores, categories, and treatment status from the live database.

Tools calledlist_risks, get_risk_dashboard
Tokens3,999 input → 948 output
ModelHaiku 4.5
Cost$0.007

Query

Which controls are failing and what risks do they affect?

Cross-references control assessments with risk register. Maps 3 failing controls to 5 affected risk scenarios.

Tools calledget_gap_analysis, list_risks, get_risk_heatmap
Tokens~8,500 input → ~2,200 output
ModelHaiku 4.5
Cost$0.02

Query

Give me a comprehensive review of our security posture

6 agents deliberate in parallel. CISO synthesises. Returns structured report with consensus, dissents, and prioritised actions.

Tools called32 tool calls across 5 council members
Tokens89,431 input → 30,539 output
ModelHaiku 4.5 (Council)
Cost$0.19

AI Agents Council

Six specialists. One synthesised answer.

Complex questions that span multiple domains trigger the AI council. Each specialist agent analyses the question from their domain perspective. The orchestrator synthesises a unified response while preserving individual reasoning for audit.

Risk Analyst

Risk scoring, scenarios, KRIs, tolerance

Controls Auditor

Control effectiveness, testing, SoA coverage

Compliance Officer

Framework alignment, regulatory requirements

Incident Commander

Incident trends, response quality, MTTR

Evidence Auditor

Evidence completeness, freshness, coverage

CISO Strategist

Executive synthesis, strategic recommendations

Example council scenarios

Should we accept the residual risk on our cloud infrastructure controls?

Risk AnalystControls AuditorCISO Strategist

Cross-domain analysis weighing control effectiveness data, incident history, and strategic risk appetite.

What is the compliance impact of the 3 open nonconformities?

Compliance OfficerControls AuditorEvidence Auditor

Maps nonconformities to framework requirements, assesses evidence gaps, and estimates remediation priority.

Prepare a board-ready summary of our security posture.

CISO StrategistRisk AnalystIncident Commander

Synthesises risk scores, incident trends, control metrics, and treatment progress into executive narrative.

Run these workflows yourself.

All four workflows and the AI council ship in the community edition. Clone the repo, start the stack, and trigger them from the gateway.