All Classes and Interfaces
Class
Description
Normalized token usage and cost statistics extracted from AI agent JSONL logs.
Build step that runs AI coding agents and can be used from freestyle and pipeline jobs.
Shared execution settings used by the AI agent builder step and command execution flow.
Per-execution customizations contributed by an
AiAgentTypeHandler, including extra
environment variables and cleanup hooks.Strategy interface for classifying a JSON log line emitted by an AI agent into a
AiAgentLogParser.ParsedLine.Parses JSONL output from AI agents into classified
AiAgentLogParser.EventView objects.Represents a single conversation event for rendering in the UI.
Per-build action that stores AI agent invocation metadata and provides the inline conversation
view on the build page.
Serialized metadata for one invocation in a run.
Strategy interface for extracting usage statistics from a single JSON log line emitted by an AI
agent.
Describable extension point for an AI agent implementation.
Format-specific log classification for Claude Code stream-json output.
Stats extractor for Claude Code / Gemini CLI JSONL output.
Format-specific log classification for Codex CLI JSONL output.
Stats extractor for Codex CLI JSONL output.
Format-specific log classification for Cursor Agent stream-json output.
Stats extractor for Cursor Agent.
In-memory registry of live AI agent builds, used to coordinate tool-call approval gates between
the running agent process and the Jenkins web UI.
Result of a user approval decision.
Live mutable state for one running build.
Immutable view model for one outstanding approval request.
Shared utility methods used by agent-specific log format classes.
Format-specific log classification for OpenCode JSONL output.
Stats extractor for OpenCode JSONL output.