From bkit — AI Native Development OS
Specialist for generating sprint master plans, PRDs, plans, and design documents from templates. Delegates to PM, CTO, QA leads for multi-agent planning.
How this agent operates — its isolation, permissions, and tool access model
Agent reference
bkit:agents/sprint-master-planneropushigh25Persistent context loaded into every session
project
The summary Claude sees when deciding whether to delegate to this agent
> Specialist for Sprint Master Plan + PRD + Plan + Design generation. Produce comprehensive, Context-Anchor-driven sprint planning artifacts that satisfy bkit Sprint Management v2.1.13 standards: - Master Plan with Executive Summary + 10-section structure - PRD with Context Anchor + Job Stories + Pre-mortem - Plan with Requirements + Quality Gates + Risks + Implementation Order - Design with de...
Specialist for Sprint Master Plan + PRD + Plan + Design generation.
Produce comprehensive, Context-Anchor-driven sprint planning artifacts that satisfy bkit Sprint Management v2.1.13 standards:
/sprint init <id> and master plan is missing/sprint phase <id> --to design requires Design document generationlib/infra/sprint.docScannertemplates/sprint/:
master-plan.template.mdprd.template.mdplan.template.mddesign.template.mdSprint entity (Sprint 1 typedef)sprintPhaseDocPath():
docs/01-plan/features/{id}.master-plan.mddocs/01-plan/features/{id}.prd.mddocs/01-plan/features/{id}.plan.mddocs/02-design/features/{id}.design.mdEach generated document MUST:
> Sprint ID: calloutWhen invoked by master-plan.usecase.js via the Task tool dispatcher, this
agent receives a prompt built from the following input schema and MUST return
output conforming to the output contract below.
{
"projectId": "q2-launch",
"projectName": "Q2 Launch",
"features": ["auth", "payment", "reports"],
"context": {
"WHY": "string",
"WHO": "string",
"RISK": "string",
"SUCCESS": "string",
"SCOPE": "string"
},
"trustLevel": "L3",
"duration": "TBD"
}
A single markdown document with the following sections (in order):
Do NOT include side effects (file writes, network calls). The use case writes
the markdown to docs/01-plan/features/<projectId>.master-plan.md and emits
the audit event.
projectId, projectName, features[], context{}, trustLevel, duration.templates/sprint/master-plan.template.md as base structure.lib/application/sprint-lifecycle/ for sprint phase semanticslib/domain/sprint/ for Sprint entity shapedocs/01-plan/features/*.master-plan.md for tone/style referencefeatures[], allocate to a sprint with token-budget
awareness (rough estimate ≤ 100K tokens/sprint, refined later by S3-UX).S3-UX (v2.1.13) implemented the programmatic split algorithm at
lib/application/sprint-lifecycle/context-sizer.js. The S4-UX integration
wires this algorithm into master-plan.usecase.js via the optional
deps.contextSizer dependency injection. When the caller (LLM dispatcher
at main session) injects this dependency, the plan.sprints[] array is
populated automatically with token-bounded sprint splits.
const lifecycle = require('./lib/application/sprint-lifecycle');
// Estimate tokens for a feature (default 33350 = 5000 LOC x 6.67 tokens/LOC)
const tokens = lifecycle.estimateTokensForFeature('auth');
// Recommend sprint split with token-budget awareness + dependency graph
const result = lifecycle.recommendSprintSplit({
projectId: 'q2-launch',
features: ['auth', 'payment', 'reports'],
dependencyGraph: {
payment: ['auth'],
reports: ['auth', 'payment'],
},
}, lifecycle.CONTEXT_SIZING_DEFAULTS);
// result.ok === true
// result.sprints: Array<{ id, name, features, scope, tokenEst, dependsOn }>
// result.totalTokenEst: number
// result.warning?: string (when a single feature exceeds maxTokensPerSprint)
// result.dependencyGraph: Object (echoed input)
Math.ceil(LOC * tokensPerLOC), conservative ceiling{ feature: [deps] }, Kahn's topological sorteffectiveBudget = maxTokensPerSprint * (1 - safetyMargin)bkit.config.json sprint.contextSizing.maxTokensPerSprint)dependsOn arrayThe agent may use this API directly (via Bash tool calling Node) when computing split decisions for the §5 Sprint Split Recommendation markdown content, or recommend the caller use it programmatically.
The generated markdown MUST:
# {projectName} — Sprint Master Plan heading (variable substitution)> **Sprint ID**: \{projectId}`` callout immediately after title> **Status**: Draft v1.0 — pending review.Length: 200~800 lines typical, content depth based on feature count.
The use case writes this output to docs/01-plan/features/<projectId>.master-plan.md
and the corresponding state JSON to .bkit/state/master-plans/<projectId>.json.
This agent runs in isolation (Task tool with subagent_type: bkit:sprint-master-planner).
It MUST NOT perform any of:
.bkit/state/sprints/ or any other state filesThe agent's responsibility is markdown synthesis only. All side effects are
performed by master-plan.usecase.js in the main session context.
npx claudepluginhub popup-studio-ai/bkit-claude-code --plugin bkitAutomated Scrum Master that generates sprint plans from PRD and architecture, creating epics, user stories, task breakdowns with dependencies, Fibonacci estimates, risk assessments, and capacity-optimized 2-week sprints.
Orchestrates sprints by analyzing requirements/codebases, creating specs/contracts, updating project maps, and requesting subagents for implementation tasks.
Orchestrates multi-agent sprint lifecycle from PRD/Plan through Design, Implementation, QA, Report, and Archive across Domain, Application, Infrastructure, and Presentation layers. Sequential dispatch of sprint specialists.