estimate

Estimate

Estimation in mobile work is risky because framework knowledge gaps, platform divergence, and store review cycles inflate timelines. This skill produces estimates with explicit reasoning and confidence levels, not a single number.

Read-only.

---

Purpose / When to Run

Run when:

• Sizing a story before adding to a sprint
• Sizing an epic before committing to a milestone
• A new feature request needs a rough budget before deciding go / no-go

Inputs

• A story path, epic slug, or glob of stories
• Recent sprint plans (for velocity baseline)
• Story files in scope (for complexity analysis)

Outputs

• Printed estimate with low / mid / high bounds, confidence, and assumptions

---

Phase 1: Resolve Scope

Argument forms:

• story <path> — one story
• epic <slug> — all stories under that epic
• batch <glob> — explicit list

If no argument, ask:

• Prompt: "Estimate what?"
• Options: Single story, Whole epic, A list of stories, A new feature with no stories yet

For "feature with no stories yet", switch to T-shirt mode (Phase 4 alternate).

---

Phase 2: Establish Velocity Baseline

Read the last 2-3 sprint plans. From each:

• Stories planned vs. completed
• Capacity in dev-days
• Average story-days

Compute the rolling average story-day cost. If sprints don't exist yet, use the default heuristic:

• Logic story: 0.5-1 day
• Integration story: 1-2 days
• UI story: 1-2 days
• Visual / Feel story: 1-3 days
• Platform story (push, biometric, IAP, deep links): 2-4 days
• Config / Data story: 0.25-0.5 day

Document the baseline source ("Computed from sprints 001-003" or "Heuristic — no sprint history").

---

Phase 3: Per-Story Complexity Analysis

For each story in scope, read it and score:

Complexity factors (each adds time)

• Type:
  • Logic: base
  • Integration: +50%
  • UI: base or +25% if novel
  • Visual: +25-50%
  • Platform: +50-100%
  • Config: -50%
• Framework risk (from story header):
  • LOW: 0
  • MEDIUM: +25%
  • HIGH: +50% (verification work)
• ADR status:
  • Accepted: 0
  • Proposed: +100% (risk that decision changes mid-implementation, or block)
• UX spec maturity (UI stories):
  • Reviewed: 0
  • Authored not reviewed: +25%
  • Missing: BLOCKED — can't estimate
• Acceptance criteria count:
  • 1-3: 0
  • 4-6: +25%
  • 7+: +50% (story may be too big — recommend split)
• Cross-platform: if both iOS and Android target → +25-50%
• Permission flow: any story that asks for permissions → +0.25 day for denied-state UX
• Backend integration: if real backend not yet ready → +50% for stub/mock work
• Store-review-affecting: any IAP, deletion, or privacy change → +0.5 day for review-prep tasks