mk:planning-engine

What This Skill Does

Pre-sprint analysis engine. Takes a set of tickets and produces three reports: technical review (complexity, risks, unknowns per ticket), capacity estimate (story points with bin-packing), and dependency graph (execution order, cycles, critical path). Delegates deep analysis to subagents (tech-analyzer, planning-reporter).

When to Use

• Sprint planning — need to estimate tickets before the sprint starts
• Backlog grooming — need technical assessment of incoming tickets
• Dependency analysis — need to understand execution order across tickets
• NOT for: creating Jira tickets (use mk:jira), assigning work, modifying sprints, breaking dependency cycles

Core Capabilities

• Technical review: 7-dimension assessment per ticket (complexity, risk, unknowns, dependencies, test surface, security surface, data impact) with 3 confidence levels
• Capacity estimation: story point estimation with bin-packing algorithm. Flags [INCOMPLETE] when >30% of tickets are unestimated
• Dependency graphing: cycle detection (reports cycles but does NOT auto-break), orphan detection, critical path computation
• Subagents: tech-analyzer (per-ticket deep analysis) and planning-reporter (report generation)
• Graceful degradation: each optional input (scout report, dependency graph, evaluate/estimate output) has a specific flag when missing

Arguments

Flag	Effect
(no flag)	Full analysis — all three reports
--tech-only	Technical review only
--capacity-only	Capacity estimate only
--deps-only	Dependency graph only

Workflow

1. Scout — gather codebase context (optional, flagged if missing)
2. Tech Review — spawn tech-analyzer subagent per ticket with injection defense (ticket content wrapped in boundary markers)
3. Capacity — run capacity-bin.py with estimated points, flag incomplete estimates
4. Dependencies — run dep-graph.py, detect cycles and orphans, compute critical path
5. Report — spawn planning-reporter subagent to synthesize findings
6. Output — produce three structured reports with templates

Usage

/mk:planning-engine PROJ-123 PROJ-124 PROJ-125
/mk:planning-engine --capacity-only sprint-42-tickets.txt

Example Prompt

Analyze these tickets for next sprint: PROJ-101 (add payment gateway), PROJ-102 (fix login timeout), PROJ-103 (migrate user profiles to new schema). I need tech review, capacity, and dependency order.

Common Use Cases

• Sprint planning with Jira ticket keys
• Backlog technical triage
• Capacity planning for team allocation
• Dependency-aware execution ordering

Pro Tips

• Always verify PTO/ceremonies/focus factor. Capacity estimation assumes 100% availability unless told otherwise.
• Cycles are surfaced, not broken. If two tickets depend on each other, the engine reports the cycle. Your team decides how to break it.
• More than 30% unestimated tickets triggers [INCOMPLETE]. Run estimates first (mk:jira estimate) before full planning.

Canonical source: .claude/skills/planning-engine/SKILL.md