AutoChecklist

A library of composable pipelines for generating and scoring checklist criteria.

A checklist is a list of yes/no questions used. autochecklist provides 5 generator abstractions, each representing a different reasoning approach to producing evaluation criteria, along with a configurable ChecklistScorer that consolidates three scoring strategies from literature. You can mix, extend, and customize all components.

Terminology

• input: The instruction, query, or task given to the LLM being evaluated (e.g., "Write a haiku about autumn").
• target: The output being evaluated against the checklist (e.g., the haiku the LLM produced).
• reference: An optional gold-standard response used by some methods to improve checklist generation.

Generator Abstractions

The core of the library is 5 generator classes, each implementing a distinct approach to producing checklists:

Instance-level generators produce one checklist per input — criteria are tailored to each specific task. Corpus-level generators produce one checklist for an entire dataset — criteria capture general quality patterns derived from higher-level signals.

Each generator is customizable via prompt templates (.md files with {input}, {target} placeholders). You can use the built-in paper implementations, write your own prompts, or chain generators with different refiners and scorers to build custom evaluation pipelines.

Built-in Pipelines

The library includes built-in pipelines implementing methods from research papers (TICK, RocketEval, RLCF, OpenRubrics, CheckEval, InteractEval, and more). See Supported Pipelines for the full list and configuration details.

Scoring

A single configurable ChecklistScorer class supports all scoring modes:

Refiners

Refiners are pipeline stages that clean up raw checklists before scoring. They're used by corpus-level generators internally, and can also be composed into custom pipelines:

• Deduplicator — merges semantically similar items via embeddings
• Tagger — filters by applicability and specificity
• UnitTester — validates that items are enforceable
• Selector — picks a diverse subset via beam search

Installation

pip install autochecklist

Optional extras

# ML dependencies for corpus-level refiners (embeddings, deduplication)
pip install "autochecklist[ml]"

# vLLM for offline GPU inference (no server needed)
pip install "autochecklist[vllm]"

# Everything
pip install "autochecklist[all]"

For development installation from source, see the GitHub repository.

Quick Start

from autochecklist import pipeline

pipe = pipeline("tick", generator_model="openai/gpt-5-mini", scorer_model="openai/gpt-5-mini")
result = pipe(input="Write a haiku about autumn.", target="Leaves fall gently down...")
print(f"Pass rate: {result.pass_rate:.0%}")

See the Quick Start guide for custom prompts, batch evaluation, and more.

CLI

autochecklist run --pipeline tick --data eval_data.jsonl -o results.jsonl \
  --generator-model openai/gpt-4o-mini --scorer-model openai/gpt-4o-mini

See the CLI guide for all commands.

Links

• Documentation
• GitHub Repository — contributing, UI, dev setup
• Bug Tracker

Citation

TBA

License

Apache-2.0 (see LICENSE)