Skip to main content

Module run

Module run 

Source
Expand description

AI Run Definition

A Run represents a single execution instance of an AI agent attempting to perform a Task. It captures the execution context (environment, agent role) and tracks the progress.

§Relationship to Task

A Task can have multiple Runs. This happens when:

  • An agent fails and retries.
  • A user requests a different approach.
  • Multiple agents work on the same task in parallel (future).

§Key Fields

  • task_id: Links back to the parent Task.
  • base_commit_sha: The Git commit hash where this run started.
  • context_snapshot_id: Links to the captured context (files, docs) used.

Structs§

AgentInstance
Agent instance participating in a run.
Environment
Environment snapshot of the run host. Captured at run creation time.
Run
Run object for a single orchestration execution. Links a task to execution state and environment.

Enums§

RunStatus
Run lifecycle status.