human-readable-workflow-name

name	<Human-readable workflow name>
slug	<workflow_slug>
version	0.1.0
status	draft
category	residential_multifamily
subsystem	residential_multifamily
pack_type	workflow
targets	["claude_code"]
stale_data	<What might drift.>
applies_to	{"segment":["middle_market"],"form_factor":[],"lifecycle":["stabilized"],"management_mode":["self_managed","third_party_managed"],"role":[],"output_types":["<pick from taxonomy Axis 9>"],"decision_severity_max":"action_requires_approval"}
references	{"reads":["reference/normalized/..."],"writes":[]}
metrics_used	["<canonical_slugs>"]
escalation_paths	[]
approvals_required	[]
description	One paragraph: what the workflow does and when it triggers.

Workflow purpose

One paragraph.

Trigger conditions

• Explicit phrases.
• Implicit signals.
• Recurring contexts.

Inputs (required / optional)

Input	Type	Required	Notes

Outputs

Output	Type	Shape

Required context

What axes / references the router must supply before the workflow can run.

Process

1. Step one.
2. Step two.

Include decision points and branches.

Metrics used

Cite by slug.

Reference files used

Paths.

Escalation points

Where the workflow hands off mid-process.

Required approvals

Threshold-triggered and policy-triggered.

Failure modes

Known ways the workflow gets wrong.

Edge cases

Unusual inputs and how the workflow handles them.

Example invocations

Example outputs