Skip to content

Tools and Templates

Put it into practice › Tools and templates

The framework is accompanied by artifacts that help put it into practice: intake forms for use case requests, machine-readable schemas for detection and use case records, a reference implementation with validation and tests, and a self-assessment instrument. This page describes what each is for and where to find it.

All of them are licensed Apache-2.0, like the rest of the repository, so they can be copied into other repositories and internal systems without licence friction.

What is here

Path Phase Purpose
use-case-request-template.md Planning Human-readable use case request
servicenow-use-case-request-form.txt Planning ServiceNow catalog item definition
salesforce-use-case-request-form.txt Planning Salesforce form definition
microsoft-forms-use-case-request-form.txt Planning Microsoft Forms question set
google-forms-use-case-request-form.txt Planning Google Forms question set

Schemas instead of templates

For anything downstream of planning, the schemas and reference implementation are the templates. They are machine-readable, validated in CI, and therefore cannot drift from the specification the way a Markdown template can.

Instead of a template for... Use
Use case request schema/use-case.schema.json and UC-2026-0001.yml
Detection documentation schema/detection.schema.json and DET-2026-0001.yml
Rule testing fixtures/DET-2026-0001/ and def_test.py
Response playbook PB-0003-oauth-consent-abuse.md
Runbook RB-0007-revoke-service-principal.md
Metrics and tracking detection-metrics.md — definitions and thresholds, not a spreadsheet
Maturity assessment assessment/

This is deliberate. A Markdown template that duplicates a schema will fall out of sync with it, and the copy people actually use will be the stale one.

Using the intake forms

The .txt files are field definitions for building the request form in an existing intake system. Field names can be adapted to the local instance.

Whichever system is used, the form MUST capture enough to satisfy schema/use-case.schema.json, because the submitted request becomes the UC- record that every resulting detection traces back to (GOV-1).

The fields most often omitted, and most often regretted:

Field Requirement Why it matters
Out of scope PLN-6 Undocumented non-goals become handover disputes
Success criteria PLN-7 "Improved visibility" cannot be verified at delivery
Business driver reference GOV-1 Without it the detection cannot be justified or defended at review
Compliance deadline PLN-4 Compliance requests are scheduled to a date, not ranked by score

Prioritisation

The request form should capture the five rubric dimensions from the planning phase: threat relevance, asset criticality, coverage gap, build cost and maintenance burden.

Requesters should not score their own requests. They supply the evidence; detection engineering and threat intelligence apply the anchored descriptors. Self-scoring produces a backlog in which every request is a five.

AI assistant skill

The senior detection engineer skill packages the rubric, schema, review checklist, metrics and conformance requirements for use by an AI assistant. It can score requests, review and write rules, produce detection records, fixtures and playbooks, and run a conformance gap assessment. Its outputs are drafts, to be verified against real telemetry and passed through the program's normal review.

Contributing templates

New templates are welcome where a machine-readable artifact would not serve better. Before contributing, check whether the proposed template should instead be an extension to the schema. See CONTRIBUTING.md.

What comes next

The remaining chapters are reference material. The specification states every requirement in the framework in testable form, and related work compares the framework with other established work in the field.