Runway

Estimate cash runway from cost-governor data, API spend rates, and known fixed costs.

Execution

1. Gather cost data

source .venv/bin/activate
python3 -c "
from hummbl_governance.integrations.cost_tracker import CostTracker
ct = CostTracker()
print('Current spend:', ct.get_current_spend())
print('Budget:', ct.get_budget())
print('Daily rate:', ct.get_daily_rate())
" 2>/dev/null || echo "(cost tracker unavailable)"

2. Known fixed costs

Item	Monthly Cost	Notes
Claude Max (the owner)	$100	Anthropic subscription
Claude Max (team)	$100	Anthropic subscription
GitHub Pro	$4	your-org org
Tailscale	$0	Free tier
Domain (your-domain.com)	~$1	Annual, amortized
GCP credits	-$2,000	2 projects x $1K (expiring)

3. Variable costs

• API calls (OpenAI, Anthropic, Google) -- from cost-governor
• Compute ($REMOTE_HOST electricity) -- estimate ~$10/mo
• Inference (OpenRouter, if used) -- from cost-governor

4. Forecast

Monthly burn = fixed_costs + avg_variable_costs
Runway = cash_on_hand / monthly_burn

Output Format

Runway Report | <date>
═══════════════════════

Monthly burn: $<X>
  Fixed: $<Y>
  Variable: $<Z> (30-day average)

Cash on hand: $<X>
Runway: <N> months

## Trend
<increasing/decreasing/stable burn rate>

## Alerts
- <if runway < 6 months: WARNING>
- <if runway < 3 months: CRITICAL>

## Cost Optimization Opportunities
- <specific suggestions based on spend patterns>

Notes

• This skill provides estimates, not accounting. For actual financials, use proper bookkeeping.
• GCP credits have expiration dates -- factor those in.
• Variable costs spike during heavy research/inference sessions.

Skill Chains

• For factor Cloudflare Neuron consumption into API spend forecasting -> [usage-monitor] (python ~/bin/usage_monitor.py status)
• For forecast API spend using free-tier model routing data -> [reasoning-router] (models)