Config format

Scenarios can be exported and imported as .conf text files. The format is human-readable and version-control friendly.

Overview

A .conf file is divided into sections, each preceded by ---. Each section contains a different part of the model.

Comments start with #.

# My process model

---
S:Intake
T:Review
T:Approval
Q:Buffer
E:Done

---
Intake ~ exponential(rate=2), constant(value=1)
Review ~ normal(mean=5, std=1)
Approval ~ lognormal(mean=2, sigma=0.5)

---
Intake -> Review
Review -> Buffer
Buffer -> Approval : 0.8
Buffer -> Done : 0.2
Approval -> Done

---
R:Analyst $50/hour #4a90d9
R:Manager $100/day

---
Analyst(1) -> Review
Manager(1) -> Approval

Nodes section

One node per line: <prefix>:<label>

Prefix	Type
S:	Start
T:	Task
Q:	Queue
E:	End

Distributions section

One distribution per line: <label> ~ <dist>

• Task nodes: <label> ~ <distribution>
• Start nodes: <label> ~ <frequency_dist>, <severity_dist>
• Queue and End nodes have no distribution.

Distribution syntax: <name>(<param>=<value>, ...)

Distribution	Parameters
constant	value
exponential	rate
normal	mean, std
uniform	low, high
poisson	mu
lognormal	mean, sigma
gamma	shape, scale
weibull	scale, shape
triangular	low, mode, high

Connections section

One connection per line: <from> -> <to> or <from> -> <to> : <probability>

Review -> Approval
Review -> Rejected : 0.1

If probabilities are omitted, they default to 1.0. When multiple connections leave the same node, probabilities should sum to 1.

Resources section

One resource per line: R:<name> $<cost>/<period> #<hex-color>

Cost and color are optional.

R:Analyst $50/hour #4a90d9
R:Manager

Resource allocations section

One allocation per line: <resource_name>(<count>) -> <node_label>

Analyst(1) -> Review
Manager(2) -> Approval

Validation errors

If a .conf file contains errors, the import will fail with a message describing the problem — for example, an unknown distribution name or a connection referencing a node that doesn't exist.