Metadata-Version: 2.4
Name: openai-gabriel
Version: 1.1.9
Summary: LLM-based library to measure quantitative attributes on qualitative data
Author: Hemanth Asirvatham, Elliott Mokski
License:                                  Apache License
                                   Version 2.0, January 2004
                                http://www.apache.org/licenses/
        
           TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
        
           1. Definitions.
        
              "License" shall mean the terms and conditions for use, reproduction,
              and distribution as defined by Sections 1 through 9 of this document.
        
              "Licensor" shall mean the copyright owner or entity authorized by
              the copyright owner that is granting the License.
        
              "Legal Entity" shall mean the union of the acting entity and all
              other entities that control, are controlled by, or are under common
              control with that entity. For the purposes of this definition,
              "control" means (i) the power, direct or indirect, to cause the
              direction or management of such entity, whether by contract or
              otherwise, or (ii) ownership of fifty percent (50%) or more of the
              outstanding shares, or (iii) beneficial ownership of such entity.
        
              "You" (or "Your") shall mean an individual or Legal Entity
              exercising permissions granted by this License.
        
              "Source" form shall mean the preferred form for making modifications,
              including but not limited to software source code, documentation
              source, and configuration files.
        
              "Object" form shall mean any form resulting from mechanical
              transformation or translation of a Source form, including but
              not limited to compiled object code, generated documentation,
              and conversions to other media types.
        
              "Work" shall mean the work of authorship, whether in Source or
              Object form, made available under the License, as indicated by a
              copyright notice that is included in or attached to the work
              (an example is provided in the Appendix below).
        
              "Derivative Works" shall mean any work, whether in Source or Object
              form, that is based on (or derived from) the Work and for which the
              editorial revisions, annotations, elaborations, or other modifications
              represent, as a whole, an original work of authorship. For the purposes
              of this License, Derivative Works shall not include works that remain
              separable from, or merely link (or bind by name) to the interfaces of,
              the Work and Derivative Works thereof.
        
              "Contribution" shall mean any work of authorship, including
              the original version of the Work and any modifications or additions
              to that Work or Derivative Works thereof, that is intentionally
              submitted to Licensor for inclusion in the Work by the copyright owner
              or by an individual or Legal Entity authorized to submit on behalf of
              the copyright owner. For the purposes of this definition, "submitted"
              means any form of electronic, verbal, or written communication sent
              to the Licensor or its representatives, including but not limited to
              communication on electronic mailing lists, source code control systems,
              and issue tracking systems that are managed by, or on behalf of, the
              Licensor for the purpose of discussing and improving the Work, but
              excluding communication that is conspicuously marked or otherwise
              designated in writing by the copyright owner as "Not a Contribution."
        
              "Contributor" shall mean Licensor and any individual or Legal Entity
              on behalf of whom a Contribution has been received by Licensor and
              subsequently incorporated within the Work.
        
           2. Grant of Copyright License. Subject to the terms and conditions of
              this License, each Contributor hereby grants to You a perpetual,
              worldwide, non-exclusive, no-charge, royalty-free, irrevocable
              copyright license to reproduce, prepare Derivative Works of,
              publicly display, publicly perform, sublicense, and distribute the
              Work and such Derivative Works in Source or Object form.
        
           3. Grant of Patent License. Subject to the terms and conditions of
              this License, each Contributor hereby grants to You a perpetual,
              worldwide, non-exclusive, no-charge, royalty-free, irrevocable
              (except as stated in this section) patent license to make, have made,
              use, offer to sell, sell, import, and otherwise transfer the Work,
              where such license applies only to those patent claims licensable
              by such Contributor that are necessarily infringed by their
              Contribution(s) alone or by combination of their Contribution(s)
              with the Work to which such Contribution(s) was submitted. If You
              institute patent litigation against any entity (including a
              cross-claim or counterclaim in a lawsuit) alleging that the Work
              or a Contribution incorporated within the Work constitutes direct
              or contributory patent infringement, then any patent licenses
              granted to You under this License for that Work shall terminate
              as of the date such litigation is filed.
        
           4. Redistribution. You may reproduce and distribute copies of the
              Work or Derivative Works thereof in any medium, with or without
              modifications, and in Source or Object form, provided that You
              meet the following conditions:
        
              (a) You must give any other recipients of the Work or
                  Derivative Works a copy of this License; and
        
              (b) You must cause any modified files to carry prominent notices
                  stating that You changed the files; and
        
              (c) You must retain, in the Source form of any Derivative Works
                  that You distribute, all copyright, patent, trademark, and
                  attribution notices from the Source form of the Work,
                  excluding those notices that do not pertain to any part of
                  the Derivative Works; and
        
              (d) If the Work includes a "NOTICE" text file as part of its
                  distribution, then any Derivative Works that You distribute must
                  include a readable copy of the attribution notices contained
                  within such NOTICE file, excluding those notices that do not
                  pertain to any part of the Derivative Works, in at least one
                  of the following places: within a NOTICE text file distributed
                  as part of the Derivative Works; within the Source form or
                  documentation, if provided along with the Derivative Works; or,
                  within a display generated by the Derivative Works, if and
                  wherever such third-party notices normally appear. The contents
                  of the NOTICE file are for informational purposes only and
                  do not modify the License. You may add Your own attribution
                  notices within Derivative Works that You distribute, alongside
                  or as an addendum to the NOTICE text from the Work, provided
                  that such additional attribution notices cannot be construed
                  as modifying the License.
        
              You may add Your own copyright statement to Your modifications and
              may provide additional or different license terms and conditions
              for use, reproduction, or distribution of Your modifications, or
              for any such Derivative Works as a whole, provided Your use,
              reproduction, and distribution of the Work otherwise complies with
              the conditions stated in this License.
        
           5. Submission of Contributions. Unless You explicitly state otherwise,
              any Contribution intentionally submitted for inclusion in the Work
              by You to the Licensor shall be under the terms and conditions of
              this License, without any additional terms or conditions.
              Notwithstanding the above, nothing herein shall supersede or modify
              the terms of any separate license agreement you may have executed
              with Licensor regarding such Contributions.
        
           6. Trademarks. This License does not grant permission to use the trade
              names, trademarks, service marks, or product names of the Licensor,
              except as required for reasonable and customary use in describing the
              origin of the Work and reproducing the content of the NOTICE file.
        
           7. Disclaimer of Warranty. Unless required by applicable law or
              agreed to in writing, Licensor provides the Work (and each
              Contributor provides its Contributions) on an "AS IS" BASIS,
              WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
              implied, including, without limitation, any warranties or conditions
              of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
              PARTICULAR PURPOSE. You are solely responsible for determining the
              appropriateness of using or redistributing the Work and assume any
              risks associated with Your exercise of permissions under this License.
        
           8. Limitation of Liability. In no event and under no legal theory,
              whether in tort (including negligence), contract, or otherwise,
              unless required by applicable law (such as deliberate and grossly
              negligent acts) or agreed to in writing, shall any Contributor be
              liable to You for damages, including any direct, indirect, special,
              incidental, or consequential damages of any character arising as a
              result of this License or out of the use or inability to use the
              Work (including but not limited to damages for loss of goodwill,
              work stoppage, computer failure or malfunction, or any and all
              other commercial damages or losses), even if such Contributor
              has been advised of the possibility of such damages.
        
           9. Accepting Warranty or Additional Liability. While redistributing
              the Work or Derivative Works thereof, You may choose to offer,
              and charge a fee for, acceptance of support, warranty, indemnity,
              or other liability obligations and/or rights consistent with this
              License. However, in accepting such obligations, You may act only
              on Your own behalf and on Your sole responsibility, not on behalf
              of any other Contributor, and only if You agree to indemnify,
              defend, and hold each Contributor harmless for any liability
              incurred by, or claims asserted against, such Contributor by reason
              of your accepting any such warranty or additional liability.
        
           END OF TERMS AND CONDITIONS
        
           APPENDIX: How to apply the Apache License to your work.
        
              To apply the Apache License to your work, attach the following
              boilerplate notice, with the fields enclosed by brackets "[]"
              replaced with your own identifying information. (Don't include
              the brackets!)  The text should be enclosed in the appropriate
              comment syntax for the file format. We also recommend that a
              file or class name and description of purpose be included on the
              same "printed page" as the copyright notice for easier
              identification within third-party archives.
        
           Copyright [yyyy] [name of copyright owner]
        
           Licensed under the Apache License, Version 2.0 (the "License");
           you may not use this file except in compliance with the License.
           You may obtain a copy of the License at
        
               http://www.apache.org/licenses/LICENSE-2.0
        
           Unless required by applicable law or agreed to in writing, software
           distributed under the License is distributed on an "AS IS" BASIS,
           WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
           See the License for the specific language governing permissions and
           limitations under the License.
        
Project-URL: Homepage, https://github.com/openai/GABRIEL
Project-URL: Repository, https://github.com/openai/GABRIEL
Project-URL: Documentation, https://github.com/openai/GABRIEL/blob/main/README.md
Project-URL: Issues, https://github.com/openai/GABRIEL/issues
Keywords: llm,gpt,openai,ai,research,academic,econ,economics,econometrics,machine learning,science,social science,text analysis
Classifier: Development Status :: 5 - Production/Stable
Classifier: Intended Audience :: Science/Research
Classifier: License :: OSI Approved :: Apache Software License
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Topic :: Scientific/Engineering :: Information Analysis
Classifier: Topic :: Software Development :: Libraries
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
License-File: NOTICE
Requires-Dist: openai>=1.109.1
Requires-Dist: pandas>=1.0.0
Requires-Dist: numpy>=1.18.0
Requires-Dist: Jinja2>=3.0.0
Requires-Dist: tqdm>=4.40.0
Requires-Dist: aiohttp>=3.7.0
Requires-Dist: aiolimiter>=1.0.0
Requires-Dist: httpx>=0.23.0
Requires-Dist: plotly>=5.0.0
Requires-Dist: matplotlib>=3.6.0
Requires-Dist: requests>=2.0.0
Requires-Dist: scipy>=1.8.0
Requires-Dist: tiktoken>=0.2.0
Requires-Dist: pydantic>=1.10.0
Requires-Dist: pypdf>=4.0.0
Requires-Dist: python-docx>=1.1.0
Requires-Dist: statsmodels>=0.14.0
Provides-Extra: dev
Requires-Dist: build>=1.0.0; extra == "dev"
Requires-Dist: pytest>=7.0.0; extra == "dev"
Requires-Dist: pytest-asyncio>=0.21.0; extra == "dev"
Requires-Dist: pytest-cov>=4.0.0; extra == "dev"
Requires-Dist: ruff>=0.4.0; extra == "dev"
Requires-Dist: mypy>=1.0.0; extra == "dev"
Provides-Extra: analysis
Requires-Dist: tabulate>=0.9.0; extra == "analysis"
Provides-Extra: notebook
Requires-Dist: ipython>=7.0.0; extra == "notebook"
Dynamic: license-file

# GABRIEL

**GABRIEL** (Generalized Attribute Based Ratings Information Extraction Library) turns messy qualitative corpora into analysis-ready datasets with GPT. It handles prompting, batching, retries, checkpointing, and audit trails so you can treat “ask the model” workflows like any other measurement instrument. From rating rhetoric across a million speeches to matching product catalogs, you focus on the research question while GABRIEL handles the operations.

📓 **Tutorial notebook** (start here!): [GABRIEL Colab notebook](https://colab.research.google.com/drive/1RMUeAWACpViqiUMlPMMwPTKyGU-OX756?usp=sharing) — also available as `gabriel_tutorial_notebook.ipynb` in this repo if you’d like to download and run it locally.

You can install the GABRIEL Python library from [PyPI](https://pypi.org/project/openai-gabriel/) with `pip install openai-gabriel` and then `import gabriel`.

> **Compatibility note:** `max_output_tokens` is deprecated in GABRIEL. Existing
> calls remain valid, but the value is ignored and emits a `FutureWarning`;
> GABRIEL leaves the output budget to the selected model.

> **Using a coding agent?** Read [`AGENTS.md`](https://github.com/openai/GABRIEL/blob/main/AGENTS.md).
> It is the repository's shared guide to model choice, data shape, scale,
> checkpoints, and quality assurance.

Read our blog post [here](https://openai.com/index/scaling-social-science-research/), our paper with validation experiments and applied examples [here](http://www.nber.org/papers/w34834), and submit feedback / bugs / feature requests [here](https://forms.gle/RKnBskuiZ64Wt9D66).

## Table of contents

- [Why GABRIEL?](#why-gabriel)
- [What can you do with GABRIEL?](#what-can-you-do-with-gabriel)
- [Installation](#installation)
- [Using GABRIEL with coding agents](#using-gabriel-with-coding-agents)
- [Quick start](#quick-start)
- [Task highlights](#task-highlights)
- [Multimodal data and web search](#multimodal-data-and-web-search)
- [Custom prompts and model routing](#custom-prompts-and-model-routing)
- [Saving, logging, and resuming](#saving-logging-and-resuming)
- [Development and testing](#development-and-testing)
- [Citation](#citation)

## Why GABRIEL?

Most of the evidence social scientists and analysts care about lives in unstructured formats: interviews, speeches, transcripts, product photos, archival scans. Modern GPT models can judge attributes, extract facts, and reason about this material with high fidelity, but building robust pipelines is still tedious. GABRIEL provides:

- 🧠 **Human-level comprehension on demand** – express the attribute the way you would brief a human coder; GABRIEL packages the prompt, context, and retries for you.
- 📊 **Quantitative outputs** – ratings (0–100), grounded comparisons, classifications, and structured extractions return as tidy DataFrames with reproducible settings.
- ⚙️ **Operational tooling** – automatic parallelism (hundreds of concurrent calls), resumable runs, raw response logs, and helper UIs make it safe to scale to very large corpora.
- 🧰 **Extensibility** – swap instructions with `additional_instructions`, bring your own templates, or drop down to `gabriel.whatever` + custom `response_fn` for bespoke prompts while still reusing the infrastructure.

The tutorial notebook walks through these ideas step-by-step—from setting up an API key to running multimodal analyses—so skim this README, then dive into the notebook for the full guided tour.

## What can you do with GABRIEL?

### A) Measure attributes on qualitative data

| Function | Purpose & Output Scale | Example Use |
| --- | --- | --- |
| `gabriel.rate` | Asks GPT to score each text / image / audio / item on natural language attributes. Output = 0-100 rating. | Measure “populist rhetoric” in a speech; “toxicity” of tweets; “luxury” in ad images. |
| `gabriel.rank` | Pairwise comparisons between texts yields ELO-like attribute ratings. Output = grounded, relative z scores for each text. | Rank technologies by “bulkiness” or artworks by “fine brushwork”. |
| `gabriel.classify` | Classifies texts / images / audio / items on whether provided labels apply. Output = one or more classes per item. | Tag news articles, product photos, or interview clips into topical categories. |
| `gabriel.extract` | Structured fact or entity extraction. One source can return one row or expand into several extracted-entity rows. | Extract every product and its name, price, description, and color from each catalog page. |
| `gabriel.discover` | Discovers natural language features which discriminate two classes of data. | Identify what distinguishes 5 star vs. 1 star reviews or successful vs. failed startups. |

### B) Clean data

| Function | Purpose & Output Scale | Example Use |
| --- | --- | --- |
| `gabriel.merge` | Creates crosswalks. Output = merged table with GPT-matched identifiers. | Match two distinct job title directories; link patent titles to product names. |
| `gabriel.deduplicate` | Detects conceptual duplicates. Maps all duplicates to one representative term. | Collapse “F-18”, “Super Hornet Fighter Jet”, “f-18 hornet” into “F-18”. |
| `gabriel.filter` | High-throughput boolean screening. Outputs items which meet natural language condition. | Subset 18M Wikipedia titles to only technologies. |
| `gabriel.deidentify` | Replaces PII with realistic, consistent fake PII. Outputs anonymized text + mapping. | Replace names, employers, addresses before sharing interview corpora. |

### C) Helper tools

| Function | Purpose & Output Scale | Example Use |
| --- | --- | --- |
| `gabriel.codify` | Passage coding: highlights snippets in text that match qualitative codes. | Flag sentences about “economic insecurity” in speeches; “stressors” mentioned in interview. |
| `gabriel.compare` | Identifies similarities / differences between paired items. Output = list of differences. | Contrast op-eds from different districts; compare two ad campaigns. |
| `gabriel.bucket` | Proposes a compact taxonomy from many terms. Output = bucket names and definitions, not row-level assignments. | Develop candidate categories for technologies, artworks, or HR complaints, then review and apply them downstream. |
| `gabriel.seed` | Enforces a representative distribution / diversity of seeds. | Initialize unique personas that match US population distribution. |
| `gabriel.poll` | Seeds personas, expands them into full biographies, and surveys them. | Simulate a synthetic opinion poll on policy, values, and open-ended attitudes. |
| `gabriel.ideate` | Generates many novel scientific theories and filters the cream of the crop. | Procure novel theories on inflation for potential research. |
| `gabriel.debias` | Post-process measurements to remove inference bias. | Ensure GPT isn't guessing climate opinions in speeches based on general political lean. |
| `gabriel.load` | Prepares a folder of text / image / audio files into a spreadsheet for use in GABRIEL. | Image directory converted into spreadsheet of file paths. |
| `gabriel.view` | UI to view sample texts with ratings / passage coding. | Spot-check classify / rating outputs; view coded passages. |
| `gabriel.paraphrase` | Rewrites texts consistently per instructions. | Summarize earnings call transcripts to remove company specifics. |
| `gabriel.whatever` | Run any GPT prompts, but leverage GABRIEL's parallelization / checkpointing. | Any set of prompts; slots into any pipeline. |

## Installation

```bash
pip install openai-gabriel

# or install directly from GitHub
pip install \
  --force-reinstall \
  git+https://github.com/openai/GABRIEL.git@main

# then run import gabriel
```

Before running a GABRIEL call, declare your GPT API key:

```bash
export OPENAI_API_KEY="sk-..."
```

In Colab, store the key in **Secrets** under `OPENAI_API_KEY` and read it with
`google.colab.userdata`; do not paste a real key into notebook source, chat,
screenshots, or Git. The tutorial shows the complete setup.

## Using GABRIEL with coding agents

The repository's [`AGENTS.md`](https://github.com/openai/GABRIEL/blob/main/AGENTS.md)
is a concise operating guide for coding agents. It
explains task selection, model routing, checkpointing, multimodal inputs, and QA.
Agents should also read `gabriel_tutorial_notebook.ipynb` closely before designing
a workflow; it is the broadest guide to what the library can do.
Three rules are especially important:

- **Verify model slugs live.** At this documentation's last review, the exact general
  models were `gpt-5.6-sol`, `gpt-5.6-terra`, and `gpt-5.6-luna`, and audio used
  `gpt-audio-1.5`. Model IDs change; check the
  [official model catalog](https://developers.openai.com/api/docs/models) and
  [pricing](https://developers.openai.com/api/docs/pricing), plus the models
  actually provisioned for your organization or gateway, instead of guessing or
  copying an old example. Luna is often a sensible starting point for many
  applications and large-scale work. For unusually subtle, knowledge-heavy, multi-constraint,
  or writing-sensitive work, pilot Terra and Sol on representative examples and
  use the smallest model that reliably meets the research standard.
- **Generally keep `n_parallels=650`, including for small tests.** It is a ceiling, not a fixed worker
  count. GABRIEL ramps up, reads rate-limit signals, retries transient failures,
  and lowers actual concurrency when needed. Temporary retries and a few slow
  stragglers are normal: while the run is progressing, let it finish. Raising
  the ceiling can help a large workload when provisioned limits support more
  throughput; lowering it may help a known deployment constraint or persistent
  failure. A small test does not require a lower ceiling.
- **Pilot, checkpoint, then scale.** Test a representative sample, inspect raw
  responses and output shape, and keep `reset_files=False` to resume a compatible
  run. Use a new `save_dir` when the model, prompt, fields, or labels change.

## Quick start

The tutorial notebook walks through many complete projects; here’s the minimal rating flow the notebook starts with. Paste this into Colab or a notebook cell so you can use `await` directly:

```python
import os
import pandas as pd

import gabriel

PATH = os.path.expanduser("~/Documents/gabriel_runs")
toy_data = pd.DataFrame(
    {
        "entity": [
            "turkey",
            "pumpkin pie",
            "green bean casserole",
            "cornbread",
        ]
    }
)

attributes = {
    "savory taste": "How savory the dish is",
    "sweet taste": "Dessert-like sweetness",
    "tangy taste": "Notes of tartness or acidity",
}

rate_results = await gabriel.rate(
    toy_data,
    column_name="entity",
    attributes=attributes,
    save_dir=os.path.join(PATH, "toy_rate"),
    model="gpt-5.6-luna",
    n_parallels=650,  # normally leave this default unchanged
    n_runs=1,
    modality="entity",
    reset_files=True,
)
rate_results.head()
```

The helper returns a `pandas.DataFrame` with one column per attribute and writes raw model responses + configs to `save_dir`. Running the same code in a plain Python script just requires wrapping the coroutine with `asyncio.run(...)`.

## Task highlights

The tutorial notebook covers full projects end-to-end. The list below matches its main use cases so you can jump to the right helper quickly.

### Measurement primitives
- **`gabriel.rate`** – assign 0–100 scores per attribute across text, entities, images, audio, or web-sourced context.
- **`gabriel.rank`** – pairwise tournaments that surface relative winners with grounded z-scores.
- **`gabriel.classify`** – single- or multi-label tagging with label definitions and consensus columns.
- **`gabriel.extract`** – extract typed fields or multiple entities from each passage, page, image, PDF, or audio item; one source may intentionally produce several rows.
- **`gabriel.discover`** – contrast two labeled corpora to learn discriminating features.

### Qualitative coding and review
- **`gabriel.codify`** highlights snippets that match qualitative codes and pairs with **`gabriel.view`** for UI-based auditing.
- **`gabriel.compare`** contrasts paired items (drafts, policies, campaigns) with concise differences/similarities.
- **`gabriel.bucket`** proposes bucket names and definitions from a large term universe; review that taxonomy, then use `classify` or another mapping step when row-level assignments are needed.

### Data prep and cleanup
- **`gabriel.load`** converts folders of media into spreadsheets with clean IDs and file paths.
- **`gabriel.merge`** / **`gabriel.deduplicate`** produce fuzzy joins and de-duplicated lists using embeddings plus GPT checks.
- **`gabriel.filter`** screens large candidate lists with natural-language conditions.
- **`gabriel.deidentify`** replaces PII with realistic stand-ins to protect privacy.

### Ideation and custom prompts
- **`gabriel.ideate`** and **`gabriel.seed`** generate diverse candidates before deeper measurement.
- **`gabriel.whatever`** runs bespoke prompts (with optional web search or custom `response_fn`) while reusing retries, logging, and checkpointing.

## Multimodal data and web search

Set `modality` to `text`, `entity`, `pdf`, `image`, `audio`, or `web` on any measurement helper. Pair `gabriel.load` with folders of media to build the right DataFrame, and use `web_search=True` when GPT should gather context before rating or extracting. The tutorial’s county-level example shows how to chain web search → rating → mapping in one flow.

For image understanding, pass `image_detail="low"`, `"high"`, `"original"`,
or `"auto"`; this is distinct from image-generation `quality`. Audio inputs need
an audio-capable model and are routed through Chat Completions rather than the
Responses API. At this documentation's 2026-07-13 review, use
`gpt-audio-1.5`; `gpt-audio-mini` is deprecated. Verify the current audio slug
before running an old example. Treat GABRIEL's pre-run cost estimate as rough
for audio, tool calls, and very long contexts; consult live API pricing for the
applicable modality and processing tier.

## Custom prompts and model routing

- Add clarifications with `additional_instructions` (e.g., mandate mutually exclusive labels).
- Swap in your own Jinja `template_path` while keeping retries and checkpoints.
- Drop to `gabriel.whatever` for fully custom prompts, attachments, or routing logic.
- Treat every model name in an example as a dated snapshot; verify the exact current slug and endpoint support before overriding a default.

## Saving, logging, and resuming

Each run expands `save_dir` (tilde and environment variables supported), writes
cleaned outputs, and checkpoints raw calls in task-specific files such as
`*_raw_responses.csv`. Leave `reset_files=False` to resume a compatible partial
run. When the research specification changes, use a fresh directory or pass
`reset_files=True` intentionally so old responses are not mixed with a new
prompt. Keep large generated outputs and private datasets outside the source
repository. `gabriel.view` reads results for quick spot checks, and helpers like
`gabriel.utils.mapmaker.MapMaker` can consume the same files downstream.

## Development and testing

Install development extras and run tests:

```bash
pip install -e .[dev]
pytest
```

Tests rely on the built-in dummy responses, so no API key is necessary. Linting and type checks (`ruff`, `mypy`) are included in the dev extras.

## Citation

If you use GABRIEL in your research, please cite:

- Asirvatham, H., Mokski, E., and Shleifer, A. (2026). *GPT as a Measurement Tool*. NBER Working Paper No. 34834.

- Asirvatham, H. and Mokski, E. (2026). *GABRIEL: Generalized Attribute-Based Ratings Information Extraction Library* (software). GitHub repository: https://github.com/openai/GABRIEL

## License

GABRIEL is released under the Apache 2.0 License. See [LICENSE](LICENSE).
