Build a Brightway foreground model and run uncertainty

list_projects
→ list_databases(project=...)
→ list_methods(project=..., filter=...)

Record the exact Brightway project and method tuple returned by the server.

create_database(project="...", name="tutorial_fg", overwrite=false)
→ write_activities(project="...", database="tutorial_fg", activities=[...])

Each activity contains code, name, unit, and exchanges. Internal technosphere links can use the other activity code; biosphere exchanges can be resolved by flow name plus categories.

run_lca(project="...", database="tutorial_fg", code="product_production", method=[...], amount=1.0)
→ contribution_analysis(result_id=...)
→ top_emissions(result_id=...)
→ dispose_result(result_id=...)

Keep the returned result_id for interpretation calls so they all refer to the same solved LCA.

set_uncertainty(project="...", database="tutorial_fg", code="material_production", scale=0.15, distribution="lognormal")

Supported distributions are lognormal, normal, uniform, and triangular. The distribution and scale are study assumptions and should be documented.

run_monte_carlo(project="...", database="tutorial_fg", code="product_production", method=[...], iterations=500, amount=1.0, seed=42)

Use a fixed seed when you need to reproduce the same sampled run. For longer studies, use run_monte_carlo_async and manage the returned job_id.

• Brightway project and foreground database
• Activity code and functional amount
• LCIA method
• Uncertainty distribution and scale
• Iteration count and seed
• Brightway MCP version