Formulas

The complete formula reference. See formulas.md in the repo for the canonical source of truth.

Inputs

Input	Type	Default	Range
complexity	S/M/L/XL	M	—
task_type	coding/bug-fix/investigation/infrastructure/data-migration/testing/design	coding	—
num_humans	integer	1	1-20
num_agents	integer	1	1-10
maturity	exploratory/partial/mostly-automated	partial	—
risk_coefficient	float	1.3	1.0-2.5
integration_overhead	float	0.15	0.05-0.30
domain_familiarity	float	1.0	0.8-1.5
human_fix_ratio	float	0.20	0.05-0.50
review_depth	light/standard/deep	standard	—
confidence_level	50/80/90	80	—
definition_phase	concept/requirements/design/ready	ready	—
org_size	solo-startup/growth/enterprise	solo-startup	—
model_tier	economy/standard/premium	standard	—
show_cost	boolean	false	—

Lookup Tables

Base Rounds

S:  3-8     M:  8-20     L:  20-50     XL: 50-120

Minutes per Round

Exploratory: 3-5    Partial: 2-3    Mostly-automated: 1-2

Task Type Multiplier

coding: 1.0    bug-fix: 1.3    investigation: 0.5    design: 1.2
testing: 1.3   infrastructure: 1.5   data-migration: 2.0

Agent Effectiveness

S: 0.9    M: 0.5    L: 0.35    XL: 0.3

Review Minutes (by depth × complexity)

         S     M     L     XL
light:   10    15    30    60
standard:20    30    60    120
deep:    40    60    120   240

Planning Minutes

S: 15-30    M: 30-60    L: 60-180    XL: 120-480

Confidence Multiplier (size-dependent)

        S      M      L      XL
50%:   1.0    1.0    1.0    0.75
80%:   1.8    1.4    1.4    1.5
90%:   2.9    2.1    2.0    2.2

Cone of Uncertainty Spread

concept: 2.0    requirements: 1.5    design: 1.2    ready: 1.0

Org Overhead

solo-startup: 1.0    growth: 1.15    enterprise: 1.3

Story Points (Mode-Aware)

S: 1-2 pts    M: 3-5 pts    L: 8-13 pts    XL: 20-40 pts

Usage depends on Cooperation Modes:

• Human-only: Points for sizing AND velocity
• Hybrid: Points for sizing only; hours for velocity and planning
• Agent-first: Points optional; plan by human review hours

Tokens Per Round (thousands)

                    S       M       L       XL
exploratory:        8k      15k     25k     40k
partial:            6k      12k     20k     35k
mostly-automated:   5k      10k     18k     30k

Output Token Ratio

S: 0.30    M: 0.30    L: 0.30    XL: 0.35

Model Tier Pricing (per 1M tokens, USD)

                Input       Output
economy:        $0.50       $2.50
standard:       $2.50       $12.00
premium:        $5.00       $25.00

See Token Estimation for the full Step 15 formula and per-model pricing table.

Output JSON Schema

{
  "title": "string",
  "complexity": "S | M | L | XL",
  "task_type": "string",
  "agent_rounds": { "min": 0, "max": 0 },
  "agent_time_minutes": { "min": 0, "max": 0 },
  "agent_effectiveness": 0.0,
  "human_review_minutes": 0,
  "human_planning_minutes": { "min": 0, "max": 0 },
  "human_fix_minutes": { "min": 0, "max": 0 },
  "integration_minutes": { "min": 0, "max": 0 },
  "task_type_multiplier": 0.0,
  "org_overhead": 0.0,
  "confidence_level": 80,
  "confidence_multiplier": 0.0,
  "total_minutes": { "min": 0, "max": 0 },
  "total_hours": { "min": 0.0, "max": 0.0 },
  "pert_expected_minutes": 0.0,
  "pert_sd_minutes": 0.0,
  "committed_hours": { "min": 0.0, "max": 0.0 },
  "risk_level": "low | medium | high",
  "risk_notes": "string",
  "warnings": ["string"],
  "team": { "humans": 0, "agents": 0 },
  "token_estimate": {
    "total_tokens": { "min": 0, "max": 0 },
    "input_tokens": { "min": 0, "max": 0 },
    "output_tokens": { "min": 0, "max": 0 },
    "pert_expected_tokens": 0,
    "model_tier": "economy | standard | premium",
    "cost_usd": { "min": 0.0, "max": 0.0 },
    "pert_expected_cost_usd": 0.0
  }
}

For the full formula steps (15 steps), see the canonical formulas.md.