ITrajectoryStore

Interfaces · AiDotNet.Agentic.SelfImproving

Stores captured AgentTrajectory records so the self-improving layer can replay, evaluate, and learn from past agent runs.

Known implementations: InMemoryTrajectoryStore

For Beginners

This is the logbook of everything the agents have done. Other parts of the system read from it to grade past runs and figure out how to do better next time.

Methods

Method Summary
Task ClearAsync(CancellationToken cancellationToken) Removes all stored trajectories.
Task<AgentTrajectory> GetAsync(string id, CancellationToken cancellationToken) Gets a trajectory by id, or null when not found.
Task<IReadOnlyList<AgentTrajectory>> GetAllAsync(CancellationToken cancellationToken) Returns all stored trajectories, oldest first.
Task<IReadOnlyList<AgentTrajectory>> QueryAsync(Func<AgentTrajectory, bool> predicate, CancellationToken cancellationToken) Returns the trajectories matching a predicate, oldest first.
Task<string> AddAsync(AgentTrajectory trajectory, CancellationToken cancellationToken) Adds a trajectory and returns its id.