Agentic CFO for autonomous financial analysis of AI infrastructure spend. Computes burn rate, runway, gross margin, and ROI-based budget reallocation.
Install
npx skillscat add strikersam/autonomous-ai-agency/financial-analyst Install via the SkillsCat registry.
SKILL.md
Skill: financial-analyst (Agentic CFO)
Purpose
Autonomous financial analysis for AI infrastructure costs. Ingests cost lines,
computes standard CFO metrics, and emits per-line budget recommendations.
When to Use
- After adding new cost lines to the infrastructure budget
- Monthly financial review cycles
- Before scaling any AI service that incurs cost
- When runway drops below 6 months (triggers CUT recommendations)
Components
| Class | Role |
|---|---|
CostLine |
Single budget line item (name, monthly_cost, revenue_attributed) |
FinancialMetrics |
Core metrics: burn rate, runway, gross margin |
BudgetOptimizer |
Greedy ROI-weighted budget reallocation |
FinancialAgent |
Recommendation engine: CUT / SCALE / HOLD / INVESTIGATE |
Quick Start
from agents.financial_analyst import (
CostLine, FinancialMetrics, FinancialAgent,
)
metrics = FinancialMetrics(
monthly_revenue=50_000,
monthly_costs=35_000,
cash_on_hand=200_000,
cost_lines=[
CostLine("gpu", 20_000, revenue_attributed=40_000, category="cogs"),
CostLine("api", 5_000, revenue_attributed=50_000, category="operations"),
],
)
agent = FinancialAgent()
recs = agent.assess(metrics)
for line_name, rec in recs.items():
print(f"{line_name}: {rec.value}")Decision Rules
| Condition | Recommendation |
|---|---|
| runway < 6 months | CUT lowest-ROI line |
| gross_margin < 0.40 | INVESTIGATE COGS lines |
| line.roi > 5.0 and category in {sales, R&D} | SCALE |
| line.roi < 0.5 and not strategic | CUT |
| otherwise | HOLD |
Testing
pytest tests/test_financial_analyst.py -v22 tests covering CostLine, FinancialMetrics, BudgetOptimizer, and FinancialAgent.
Branch
fix/quick-note-236-agentic-cfo