Crate shiplog_cluster_llm 

LLM-assisted workstream clustering adapters.

Provides LlmClusterer plus pluggable backends for OpenAI-compatible APIs, and LlmWithFallback for repo-based fallback on LLM failure.

Re-exports

pub use parse::parse_llm_response;

pub use prompt::chunk_events;

pub use prompt::format_event_list;

pub use prompt::summarize_event;

pub use prompt::system_prompt;

Modules

parse

prompt

Structs

FailingLlmBackend

Mock backend that always fails.

LlmClusterer

LLM-assisted workstream clusterer.

LlmConfig

Configuration for LLM-assisted clustering.

LlmWithFallback

Wrapper that falls back to repo-based clustering on LLM failure.

MockLlmBackend

Mock backend for testing.

OpenAiCompatibleBackend

Backend that speaks the OpenAI chat completions protocol.

Traits

LlmBackend

Abstraction over LLM APIs. Enables testing with mocks.