6fdac22876349021634c8a3ffda0d56b3bdd7797
The operating model is refused while an Operational requirement is verified by no Check, from it or from a Condition it comes from; the architect is told so, and the operating-chain Skill lists it. Six of nine earlier runs left a requirement unverified. The evaluation checks it too, runs each case N times (--repeat), reports the spread of people, cost, Roles, Checks and Capabilities (spread.json), and writes report.md to publish. Result folders are relative to the run. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Target operating model agent
Turns a description of a solution into a reviewed target operating model: who runs it, through which processes, to which service levels, with how many people and at what cost.
It is the Execution engine (../orchestration_engine) and one Workflow. Vocabulary is in CONTEXT.md, decisions in docs/adr/, the design in .scratch/target-operating-model/spec.md, the plan in IMPLEMENTATION_PLAN.md, and what has been built in .agents/plans/.
What it produces
An Execution starts from an operating spec and goes through five Phases, each approved before the next begins.
| Phase | Deliverable | What code checks or writes |
|---|---|---|
| Scoping | Operating scope | The spec is read against a catalogue; nothing essential may be left unknown |
| Operational requirements | Operational requirements | Each is traced to what the customer stated; everything that binds is carried |
| Operating model design | Capability model | Each Capability is from the catalogue; every Item is covered and every requirement served |
| Target operating model | One accountable Role for each Capability, someone responsible at its hours, every Condition carried by a Check. The responsibility matrix, organisation chart, staffing and cost are written from the graph | |
| Validation | Validation results | A verdict on every Check; scenario walk-throughs generated and traced by code |
| Packaging | Operating model guide |
Layout
app/operating_model/ the Workflow: names, the spec reader, the chain's checks, the
graph, staffing and cost, scenarios, and the Workflow value
app/main.py python -m app.main serves the agent over A2A
agents/ role files, one per agent
providers/gcp.yaml the first Cloud provider's bundle
skills/ operating-scoping (the spec's catalogue), operating-capabilities
(the capability catalogue and its sources), operating-chain
(the data files), rate-card (the service provider's prices)
tests/ the deterministic suite; no model and no network
Prices
skills/rate-card/assets/rates.yaml is the service provider's rate card. Its figures are placeholders while its version starts with draft, and every cost says so.
Description
Languages
Python
88.9%
HCL
6.8%
Mermaid
3.7%
Dockerfile
0.6%