graph¶

Manage requirement graphs using Graphize integration.

Overview¶

The graph command extracts requirement graphs from spec files, enabling traceability analysis and visualization. It integrates with graphize for graph operations.

Subcommands¶

extract¶

Extract a requirement graph from all specs in the project.

multispec graph extract

Output:

Creates .graphize/spec-graph.json in the project directory
Prints node count summary by type

Example:

Extracted graph with 45 nodes and 32 edges
Saved to: docs/specs/user-onboarding/.graphize/spec-graph.json

Node types:
  spec: 4
  section: 18
  requirement: 12
  user_story: 6
  constraint: 3
  decision: 2

export¶

Export the graph to various formats for visualization or analysis.

multispec graph export [--format FORMAT] [--output DIR]

Flags:

Flag	Default	Description
`--format`	`html`	Export format: `html`, `graphml`, `json`
`--output`	`.graphize`	Output directory

Formats:

Format	Output File	Description
`html`	`graph.html`	Interactive HTML visualization
`graphml`	`spec-graph.graphml`	GraphML for graph tools (yEd, Gephi)
`json`	`spec-graph.json`	Raw graph data

Examples:

# Export as interactive HTML
multispec graph export --format html

# Export as GraphML for yEd
multispec graph export --format graphml

# Export to custom directory
multispec graph export --format html --output ./reports

query¶

Query nodes in the graph by type or spec.

multispec graph query [--type TYPE] [--spec SPEC]

Flags:

Flag	Description
`--type`	Filter by node type (requirement, user_story, constraint, decision)
`--spec`	Filter by spec type (mrd, prd, uxd, trd)

Examples:

# List all requirements
multispec graph query --type requirement

# List all PRD nodes
multispec graph query --spec prd

# List constraints from TRD
multispec graph query --type constraint --spec trd

Output:

Found 12 nodes

[requirement] The system shall support SSO login
  ID: prd_req_req_prd_1
  Spec: prd
  Text: The system shall support SSO login via SAML 2.0...

[requirement] Users must be able to reset passwords
  ID: prd_req_req_prd_2
  Spec: prd
  Text: Users must be able to reset passwords via email...

Node Types¶

The graph extractor identifies the following node types from specs:

Type	Source	Pattern
`requirement`	PRD	Statements with "shall", "must", "should"
`user_story`	PRD	"As a... I want... so that..."
`acceptance_criteria`	PRD	"Given/When/Then" patterns
`constraint`	MRD, TRD	Sections titled "Constraints", "Limitations"
`decision`	TRD	Sections titled "Decisions", "Architecture"
`section`	All	Markdown headings
`spec`	All	Root node for each spec file

Edge Types¶

Relationships extracted between nodes:

Type	Description
`contains`	Section contains requirements/decisions
`traces_to`	Requirement traces to technical decision
`derived_from`	TRD derived from PRD and MRD

Graph Storage¶

Graphs are stored in the .graphize/ directory within each project:

docs/specs/{project}/
└── .graphize/
    ├── spec-graph.json      # Primary graph data
    ├── spec-graph.graphml   # GraphML export (if generated)
    └── graph.html           # HTML visualization (if generated)

Integration with Graphize¶

The graph command uses:

graphize v0.3.0 - Graph extraction and export
graphfs v0.2.0 - Graph data structures

For advanced graph analysis, you can use the graphize CLI directly on the exported JSON:

# Install graphize
go install github.com/plexusone/graphize/cmd/graphize@latest

# Analyze the spec graph
graphize analyze .graphize/spec-graph.json

# Generate detailed report
graphize report .graphize/spec-graph.json