modelagem-de-grafo-de-metas-narrativas-gmn

Create each narrative goal as a vertex with attributes like title, status, and associated chapter, then register each causal dependency as a directed edge meaning one goal must complete before another. Store vertices and edges in relational tables such as goals and goal_dependencies.

Run Kahn's topological sort algorithm over the full graph each time a new dependency edge is created. If the algorithm detects a cycle, block the edge insertion in the UI with an explanatory message before it is persisted to the database.

Breadth-first search is the default propagation algorithm, with depth-first search as an alternative. Starting from the edited goal node, the traversal applies the Inconsistent flag in cascade to all descendant vertices in the directed graph.

No. The completion validation blocks any status transition to Concluído when a parent goal still has status Pendente or Inconsistente. The UI displays a blocking message listing the unfinished predecessor goals.

Avoid this approach for short, fully linear stories without intertwined subplots, where the graph overhead adds no value. It is also unsuited for software development task management, where dedicated tools like Jira are more appropriate.

The validation criteria require acyclicity checks and full inconsistency propagation on a 200-goal DAG to complete in under 5 milliseconds on a local CPU. The default maximum dependency nesting depth is 10 levels, configurable via MAX_GOAL_DEPTH.