| name | financial-planning |
| description | Build or update a comprehensive household financial plan using integrated net-worth, cashflow, and the reduced exact household-tax surface where supported. |
| user-invocable | false |
Financial Plan
Use tool-first baselines and the reduced exact household-tax surface where supported. Fail closed when the requested planning case is outside the exact tax engine scope.
MCP Tool Map
- Cashflow reality:
actual-budget.analytics(operation="monthly_summary"), actual-budget.analytics(operation="spending_by_category")
- Investments:
ghostfolio.portfolio(operation="summary"), ghostfolio.portfolio(operation="dividends")
- Net worth and liabilities:
finance-graph.get_net_worth, finance-graph.get_liability_summary, finance-graph.list_liabilities
- Tax evidence:
paperless.search_documents, paperless.get_document
- Tax strategy engine:
household-tax.assess_exact_support, household-tax.ingest_return_facts, household-tax.compute_individual_return_exact, household-tax.compute_fiduciary_return_exact, household-tax.plan_individual_safe_harbor, household-tax.plan_fiduciary_safe_harbor, household-tax.compare_trust_distribution_strategies
Workflow
- Build household baseline (income, spend, portfolio, liabilities, net worth) from MCP tools.
- Ingest and reconcile tax-document evidence from Paperless for key assumptions.
- Assess whether the requested tax-planning slice is inside the exact 2026
US + MA support surface.
- If supported, persist canonical facts with
ingest_return_facts and run the exact return / safe-harbor / trust-distribution tools that fit the case.
- If unsupported, stop and explain the missing/unsupported facts rather than synthesizing a broad tax optimization.
- Present prioritized actions with expected tax and cashflow effects only for supported exact cases.
Output Contract
Always include:
- baseline snapshot (
as_of, net worth, cashflow, liabilities)
- exact tax tools run and support status
- recommended exact actions with tradeoffs
- assumptions + sensitivity notes
- provenance and data gaps
