AI operating model

Concretize how the organization will use AI after the first experiments: goals, human review, owners, cadence. Still vendor-neutral.

Document information

Field	Value
Canonical URL	`/docs/07_ai-agents-and-mcp/model_skills/60_ai_adoption/ai-operating-model`
Version (published date)	`2026-08-27`
Tags	`ai`, `skills`, `playbooks`, `governance`

Kind

Sequential procedure.

Depends on

ai-adoption-data-model, ai-roadmap.

Prerequisites

id	question	search	if_missing_or_ambiguous	default_to_propose	binds_to
`engagements_schema`	Engagements schema	`ai-adoption-data-model`	Import first	`ai_engagements`	Tenant `SKILL.md`
`engagement_folder`	Deliverable folder	`ai-adoption-data-model`	Import first	`engagements/current/`	Tenant `SKILL.md`

Import contract

Field	Value
`suggested_skill_id`	`ai-operating-model`
`version`	`1.0.0`
`capabilities.tools`	`query_datapool`, `request_human_input`, `write_file`

Procedure

DoD: a short operating-model.md covering goals, which decisions stay human, owners, and a weekly or biweekly review. No platform resource creates.

Tenant SKILL.md (copy this body)

# AI operating model

Write `{{engagement_folder}}/operating-model.md` for the current engagement in `{{engagements_schema}}`.

## Sequence

1. Read the engagement and `now` roadmap items.
2. Draft: outcome goals (tied to platform strategy), human-in-the-loop vs unattended, named owners, review cadence, what “stop automating” looks like.
3. `request_human_input` confirm, then `write_file`.

## Stop

- Do not create ProcessFlow or Autonomic Agents.
- Do not invent owners; ask if missing.

Optional skill.json

{
  "name": "ai-operating-model",
  "version": "1.0.0",
  "description": "Write a short vendor-neutral AI operating model for the engagement.",
  "capabilities": {
    "prompt": ["Short operating-model.md. Confirm owners. No platform creates."],
    "resources": [],
    "tools": ["query_datapool", "request_human_input", "write_file"]
  }
}