Command Reference¶

Complete reference for all prdtool commands.

init¶

Initialize a new PRD with required metadata.

prdtool init --title <title> --owner <owner> [--id <id>] [-f <file>]

Flag	Required	Description	Default
`--title`	Yes	PRD title
`--owner`	Yes	PRD owner name
`--id`	No	PRD ID	Auto-generated
`-f, --file`	No	Output file path	`PRD.json`

Examples:

prdtool init --title "User Authentication" --owner "Jane PM"
prdtool init --title "Search Feature" --owner "John PM" --id PRD-2026-042
prdtool init --title "Mobile App" --owner "Team Lead" -f mobile-prd.json

show¶

Display PRD contents as formatted JSON.

prdtool show [file] [-f <file>]

Examples:

prdtool show
prdtool show my-prd.json
prdtool show -f feature.json

validate¶

Validate a PRD file against the schema.

prdtool validate [file] [-f <file>]

Performs:

JSON schema validation
ID format checking (e.g., REQ-1, RISK-2)
Traceability verification
Required field validation

Examples:

prdtool validate
prdtool validate PRD.json

Output:

{
  "valid": true,
  "errors": [],
  "warnings": ["No metrics defined"]
}

score¶

Score a PRD's quality against the rubric.

prdtool score [file] [-f <file>] [--verbose] [--json]

Flag	Description
`-v, --verbose`	Show detailed scoring breakdown
`--json`	Output as JSON

Scoring Categories:

Category	Weight
Problem Definition	20%
Solution Fit	15%
User Understanding	10%
Market Awareness	10%
Scope Discipline	10%
Requirements Quality	10%
Metrics Quality	10%
UX Coverage	5%
Technical Feasibility	5%
Risk Management	5%

Thresholds:

≥8.0 → Approve
≥6.5 → Revise
<6.5 → Human Review
≤3.0 → Blocker

Examples:

prdtool score
prdtool score --verbose
prdtool score --json | jq '.overall_score'

view¶

Generate human-readable PRD views.

prdtool view [file] [-f <file>] [-t <type>] [-o <format>]

Flag	Description	Options	Default
`-t, --type`	View type	`pm`, `exec`	`pm`
`-o, --format`	Output format	`markdown`, `json`	`markdown`

View Types:

pm: Product Manager view - detailed operational information
exec: Executive view - high-level decision summary with scores

Examples:

prdtool view                          # PM view, markdown
prdtool view --type exec              # Executive summary
prdtool view --type pm --format json  # PM view as JSON
prdtool view -t exec -o markdown > summary.md

deploy¶

Generate AI assistant configurations.

prdtool deploy [-t <target>] [-o <output>]

Flag	Description	Options	Default
`-t, --target`	Deployment target	`kiro-cli`, `kiro-power`, `claude`, `all`	`kiro-cli`
`-o, --output`	Output directory		Platform-specific

Targets:

kiro-cli: AWS Kiro CLI (MCP config + agent configs)
kiro-power: AWS Kiro IDE Power (POWER.md + mcp.json + steering/)
claude: Claude Code (MCP config)
all: All supported platforms

Examples:

prdtool deploy --target claude
prdtool deploy --target kiro-power --output ~/.kiro/powers/prdtool
prdtool deploy --target all

add¶

Add items to various sections of a PRD. All add commands support the -f, --file flag.

add problem¶

Add a problem statement.

prdtool add problem --statement <text> [--impact <text>] [--confidence <0-1>]

Flag	Required	Description	Default
`--statement`	Yes	Problem statement
`--impact`	No	User impact description
`--confidence`	No	Confidence level (0-1)	0.5

prdtool add problem --statement "Users cannot securely access accounts" --impact "Lost revenue" --confidence 0.8

add persona¶

Add a user persona.

prdtool add persona --name <name> [--role <role>] [--pain-point <text>...]

Flag	Required	Description
`--name`	Yes	Persona name
`--role`	No	Persona role
`--pain-point`	No	Pain points (repeatable)

prdtool add persona --name "Developer Dan" --role "Backend Developer" --pain-point "Slow builds" --pain-point "Complex configs"

add goal¶

Add a goal.

prdtool add goal --statement <text>

prdtool add goal --statement "Reduce authentication time by 50%"

add nongoal¶

Add a non-goal (explicit scope exclusion).

prdtool add nongoal --statement <text>

prdtool add nongoal --statement "Mobile app support is out of scope"

add solution¶

Add a solution option.

prdtool add solution --name <name> [--description <text>] [--tradeoff <text>...]

Flag	Required	Description
`--name`	Yes	Solution name
`--description`	No	Solution description
`--tradeoff`	No	Tradeoffs (repeatable)

prdtool add solution --name "OAuth 2.0" --description "Industry standard auth" --tradeoff "Complex setup" --tradeoff "Third-party dependency"

add req¶

Add a functional requirement.

prdtool add req --description <text> [--priority <level>] [--ac <text>...]

Flag	Required	Description	Default
`--description`	Yes	Requirement description
`--priority`	No	Priority: `must`, `should`, `could`	`should`
`--ac`	No	Acceptance criteria (repeatable)

prdtool add req --description "Support OAuth 2.0 login" --priority must --ac "User can login with Google" --ac "User can login with GitHub"

add nfr¶

Add a non-functional requirement.

prdtool add nfr --requirement <text> [--category <cat>]

Flag	Required	Description	Default
`--requirement`	Yes	NFR description
`--category`	No	Category (see below)	`performance`

Categories: performance, security, reliability, scalability, usability, compliance, accessibility

prdtool add nfr --requirement "API response time < 200ms" --category performance
prdtool add nfr --requirement "All data encrypted at rest" --category security

add metric¶

Add a success metric.

prdtool add metric --name <name> [--definition <text>] [--target <text>] [--type <type>]

Flag	Required	Description	Default
`--name`	Yes	Metric name
`--definition`	No	How metric is calculated
`--target`	No	Target value
`--type`	No	Type: `northstar`, `supporting`, `guardrail`	`supporting`

prdtool add metric --name "Login Success Rate" --definition "Successful logins / Total attempts" --target "99.5%" --type northstar

add risk¶

Add a risk.

prdtool add risk --description <text> [--impact <level>] [--mitigation <text>]

Flag	Required	Description	Default
`--description`	Yes	Risk description
`--impact`	No	Impact: `low`, `medium`, `high`	`medium`
`--mitigation`	No	Mitigation strategy

prdtool add risk --description "Third-party OAuth provider outage" --impact high --mitigation "Implement fallback local auth"

add decision¶

Add a decision record.

prdtool add decision --decision <text> [--rationale <text>] [--by <name>]

Flag	Required	Description
`--decision`	Yes	Decision made
`--rationale`	No	Rationale for decision
`--by`	No	Who made the decision

prdtool add decision --decision "Use JWT for session management" --rationale "Stateless, scalable" --by "Tech Lead"