open-research-agent 0.1.0__tar.gz
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- open_research_agent-0.1.0/.github/workflows/publish.yml +31 -0
- open_research_agent-0.1.0/.gitignore +19 -0
- open_research_agent-0.1.0/CHANGELOG.md +14 -0
- open_research_agent-0.1.0/CONTRIBUTING.md +72 -0
- open_research_agent-0.1.0/LICENSE +21 -0
- open_research_agent-0.1.0/PKG-INFO +186 -0
- open_research_agent-0.1.0/README.md +150 -0
- open_research_agent-0.1.0/ora/__init__.py +2 -0
- open_research_agent-0.1.0/ora/__main__.py +6 -0
- open_research_agent-0.1.0/ora/agents/__init__.py +0 -0
- open_research_agent-0.1.0/ora/agents/researcher.py +366 -0
- open_research_agent-0.1.0/ora/agents/reviewer.py +73 -0
- open_research_agent-0.1.0/ora/agents/supervisor.py +95 -0
- open_research_agent-0.1.0/ora/agents/writer.py +128 -0
- open_research_agent-0.1.0/ora/cli.py +322 -0
- open_research_agent-0.1.0/ora/config.py +131 -0
- open_research_agent-0.1.0/ora/graph.py +93 -0
- open_research_agent-0.1.0/ora/progress.py +39 -0
- open_research_agent-0.1.0/ora/prompts/__init__.py +7 -0
- open_research_agent-0.1.0/ora/prompts/researcher.py +16 -0
- open_research_agent-0.1.0/ora/prompts/reviewer.py +46 -0
- open_research_agent-0.1.0/ora/prompts/supervisor.py +41 -0
- open_research_agent-0.1.0/ora/prompts/writer.py +33 -0
- open_research_agent-0.1.0/ora/state.py +83 -0
- open_research_agent-0.1.0/ora/tools/__init__.py +0 -0
- open_research_agent-0.1.0/ora/tools/evaluate.py +112 -0
- open_research_agent-0.1.0/ora/tools/scrape.py +37 -0
- open_research_agent-0.1.0/ora/tools/search.py +48 -0
- open_research_agent-0.1.0/pyproject.toml +61 -0
- open_research_agent-0.1.0/tests/__init__.py +0 -0
- open_research_agent-0.1.0/tests/test_cli.py +380 -0
- open_research_agent-0.1.0/tests/test_config.py +52 -0
- open_research_agent-0.1.0/tests/test_evaluate.py +63 -0
- open_research_agent-0.1.0/tests/test_graph.py +55 -0
- open_research_agent-0.1.0/tests/test_integration.py +80 -0
- open_research_agent-0.1.0/tests/test_package_metadata.py +21 -0
- open_research_agent-0.1.0/tests/test_progress.py +46 -0
- open_research_agent-0.1.0/tests/test_researcher.py +32 -0
- open_research_agent-0.1.0/tests/test_researcher_intensity.py +57 -0
- open_research_agent-0.1.0/tests/test_researcher_progress.py +245 -0
- open_research_agent-0.1.0/tests/test_reviewer.py +44 -0
- open_research_agent-0.1.0/tests/test_search.py +10 -0
- open_research_agent-0.1.0/tests/test_state.py +42 -0
- open_research_agent-0.1.0/tests/test_writer_progress.py +97 -0
|
@@ -0,0 +1,31 @@
|
|
|
1
|
+
name: Publish to PyPI
|
|
2
|
+
|
|
3
|
+
on:
|
|
4
|
+
release:
|
|
5
|
+
types: [published]
|
|
6
|
+
|
|
7
|
+
jobs:
|
|
8
|
+
build-and-publish:
|
|
9
|
+
name: Build and publish to PyPI
|
|
10
|
+
runs-on: ubuntu-latest
|
|
11
|
+
environment: pypi
|
|
12
|
+
permissions:
|
|
13
|
+
id-token: write # required for OIDC trusted publishing
|
|
14
|
+
contents: read
|
|
15
|
+
|
|
16
|
+
steps:
|
|
17
|
+
- uses: actions/checkout@v4
|
|
18
|
+
|
|
19
|
+
- name: Set up Python
|
|
20
|
+
uses: actions/setup-python@v5
|
|
21
|
+
with:
|
|
22
|
+
python-version: "3.11"
|
|
23
|
+
|
|
24
|
+
- name: Install build tools
|
|
25
|
+
run: python -m pip install --upgrade build
|
|
26
|
+
|
|
27
|
+
- name: Build package
|
|
28
|
+
run: python -m build
|
|
29
|
+
|
|
30
|
+
- name: Publish to PyPI
|
|
31
|
+
uses: pypa/gh-action-pypi-publish@release/v1
|
|
@@ -0,0 +1,19 @@
|
|
|
1
|
+
__pycache__/
|
|
2
|
+
*.pyc
|
|
3
|
+
*.pyo
|
|
4
|
+
.pytest_cache/
|
|
5
|
+
dist/
|
|
6
|
+
build/
|
|
7
|
+
*.egg-info/
|
|
8
|
+
.venv/
|
|
9
|
+
venv/
|
|
10
|
+
.env
|
|
11
|
+
.coverage
|
|
12
|
+
htmlcov/
|
|
13
|
+
.ruff_cache/
|
|
14
|
+
|
|
15
|
+
# Generated ORA research reports
|
|
16
|
+
*-20??-??-??-??????.md
|
|
17
|
+
|
|
18
|
+
# Internal development documentation
|
|
19
|
+
docs/
|
|
@@ -0,0 +1,14 @@
|
|
|
1
|
+
# Changelog
|
|
2
|
+
|
|
3
|
+
## 0.1.0
|
|
4
|
+
|
|
5
|
+
Initial public release of ORA.
|
|
6
|
+
|
|
7
|
+
- Multi-agent research CLI.
|
|
8
|
+
- PyPI package name `open-research-agent`, with `open-research-agent` as the primary CLI command and `ora` as an alias.
|
|
9
|
+
- Research planning flow.
|
|
10
|
+
- Configurable intensity levels 1-5.
|
|
11
|
+
- Firecrawl search and scrape integration.
|
|
12
|
+
- DeepSeek-backed planning, research, writing, and review agents.
|
|
13
|
+
- Optional adversarial review for higher-intensity research.
|
|
14
|
+
- Local markdown report output.
|
|
@@ -0,0 +1,72 @@
|
|
|
1
|
+
# Contributing
|
|
2
|
+
|
|
3
|
+
Thanks for your interest in improving Open Research Agent (ORA).
|
|
4
|
+
|
|
5
|
+
## Development setup
|
|
6
|
+
|
|
7
|
+
Use Python 3.10 or newer.
|
|
8
|
+
|
|
9
|
+
```bash
|
|
10
|
+
python3 -m venv .venv
|
|
11
|
+
source .venv/bin/activate
|
|
12
|
+
pip install -e ".[dev]"
|
|
13
|
+
```
|
|
14
|
+
|
|
15
|
+
## Configuration for local runs
|
|
16
|
+
|
|
17
|
+
ORA 0.1.0 currently supports the DeepSeek API as its LLM backend and Firecrawl for search and scraping.
|
|
18
|
+
|
|
19
|
+
```bash
|
|
20
|
+
export DEEPSEEK_API_KEY="your-deepseek-api-key"
|
|
21
|
+
export FIRECRAWL_API_KEY="your-firecrawl-api-key"
|
|
22
|
+
```
|
|
23
|
+
|
|
24
|
+
You can create a local config file with:
|
|
25
|
+
|
|
26
|
+
```bash
|
|
27
|
+
open-research-agent config --init
|
|
28
|
+
```
|
|
29
|
+
|
|
30
|
+
## Run tests
|
|
31
|
+
|
|
32
|
+
```bash
|
|
33
|
+
pytest
|
|
34
|
+
```
|
|
35
|
+
|
|
36
|
+
For CLI-focused changes:
|
|
37
|
+
|
|
38
|
+
```bash
|
|
39
|
+
python3 -m pytest tests/test_cli.py -q
|
|
40
|
+
```
|
|
41
|
+
|
|
42
|
+
## Run the CLI locally
|
|
43
|
+
|
|
44
|
+
```bash
|
|
45
|
+
open-research-agent --help
|
|
46
|
+
ora --help
|
|
47
|
+
python -m ora --help
|
|
48
|
+
open-research-agent config --show
|
|
49
|
+
```
|
|
50
|
+
|
|
51
|
+
## Generated reports
|
|
52
|
+
|
|
53
|
+
`open-research-agent research` can generate timestamped markdown reports in the repository root or current working directory. Do not commit generated research reports.
|
|
54
|
+
|
|
55
|
+
Use `--no-save` while testing if you do not need a report file:
|
|
56
|
+
|
|
57
|
+
```bash
|
|
58
|
+
open-research-agent research "AI memory systems" --no-save
|
|
59
|
+
```
|
|
60
|
+
|
|
61
|
+
## Secrets
|
|
62
|
+
|
|
63
|
+
Do not commit API keys, `.env` files, credentials, or generated files containing private research output.
|
|
64
|
+
|
|
65
|
+
## Pull requests
|
|
66
|
+
|
|
67
|
+
Before opening a pull request:
|
|
68
|
+
|
|
69
|
+
1. Run the relevant focused tests.
|
|
70
|
+
2. Run the full test suite with `pytest`.
|
|
71
|
+
3. Check `git status --short` for generated reports or other accidental files.
|
|
72
|
+
4. Update documentation when behavior changes.
|
|
@@ -0,0 +1,21 @@
|
|
|
1
|
+
MIT License
|
|
2
|
+
|
|
3
|
+
Copyright (c) 2026 Cameron Palmer
|
|
4
|
+
|
|
5
|
+
Permission is hereby granted, free of charge, to any person obtaining a copy
|
|
6
|
+
of this software and associated documentation files (the "Software"), to deal
|
|
7
|
+
in the Software without restriction, including without limitation the rights
|
|
8
|
+
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
|
9
|
+
copies of the Software, and to permit persons to whom the Software is
|
|
10
|
+
furnished to do so, subject to the following conditions:
|
|
11
|
+
|
|
12
|
+
The above copyright notice and this permission notice shall be included in all
|
|
13
|
+
copies or substantial portions of the Software.
|
|
14
|
+
|
|
15
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
16
|
+
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
17
|
+
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
|
18
|
+
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|
19
|
+
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
|
20
|
+
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
|
21
|
+
SOFTWARE.
|
|
@@ -0,0 +1,186 @@
|
|
|
1
|
+
Metadata-Version: 2.4
|
|
2
|
+
Name: open-research-agent
|
|
3
|
+
Version: 0.1.0
|
|
4
|
+
Summary: Open Research Agent - multi-agent research with adversarial review
|
|
5
|
+
Project-URL: Homepage, https://github.com/cameronmpalmer/open-research-agent
|
|
6
|
+
Project-URL: Repository, https://github.com/cameronmpalmer/open-research-agent
|
|
7
|
+
Project-URL: Issues, https://github.com/cameronmpalmer/open-research-agent/issues
|
|
8
|
+
License: MIT
|
|
9
|
+
License-File: LICENSE
|
|
10
|
+
Keywords: deepseek,firecrawl,langchain,langgraph,multi-agent,research
|
|
11
|
+
Classifier: Environment :: Console
|
|
12
|
+
Classifier: Intended Audience :: Developers
|
|
13
|
+
Classifier: License :: OSI Approved :: MIT License
|
|
14
|
+
Classifier: Programming Language :: Python :: 3
|
|
15
|
+
Classifier: Programming Language :: Python :: 3.10
|
|
16
|
+
Classifier: Programming Language :: Python :: 3.11
|
|
17
|
+
Classifier: Programming Language :: Python :: 3.12
|
|
18
|
+
Classifier: Topic :: Scientific/Engineering :: Information Analysis
|
|
19
|
+
Requires-Python: >=3.10
|
|
20
|
+
Requires-Dist: click>=8.0
|
|
21
|
+
Requires-Dist: firecrawl-py>=1.0
|
|
22
|
+
Requires-Dist: langchain-core>=0.3.0
|
|
23
|
+
Requires-Dist: langchain-openai>=0.3.0
|
|
24
|
+
Requires-Dist: langchain>=0.3.0
|
|
25
|
+
Requires-Dist: langgraph>=0.3.0
|
|
26
|
+
Requires-Dist: pydantic-settings>=2.0
|
|
27
|
+
Requires-Dist: pydantic>=2.0
|
|
28
|
+
Requires-Dist: python-dotenv>=1.0
|
|
29
|
+
Requires-Dist: pyyaml>=6.0
|
|
30
|
+
Requires-Dist: rich>=13.0
|
|
31
|
+
Provides-Extra: dev
|
|
32
|
+
Requires-Dist: pytest-asyncio>=0.24; extra == 'dev'
|
|
33
|
+
Requires-Dist: pytest-mock>=3.14; extra == 'dev'
|
|
34
|
+
Requires-Dist: pytest>=8.0; extra == 'dev'
|
|
35
|
+
Description-Content-Type: text/markdown
|
|
36
|
+
|
|
37
|
+
# Open Research Agent (ORA)
|
|
38
|
+
|
|
39
|
+
Open Research Agent (ORA) is an open-source multi-agent research CLI. ORA plans research, searches and scrapes web sources, synthesizes findings, and optionally uses an adversarial reviewer for higher-intensity research.
|
|
40
|
+
|
|
41
|
+
Current release: **0.1.0**
|
|
42
|
+
|
|
43
|
+
## What ORA does
|
|
44
|
+
|
|
45
|
+
ORA turns a research question into a sourced markdown report:
|
|
46
|
+
|
|
47
|
+
1. A supervisor drafts a research plan.
|
|
48
|
+
2. The researcher searches and scrapes web sources.
|
|
49
|
+
3. The writer synthesizes findings into a report.
|
|
50
|
+
4. For intensity levels 4 and 5, an adversarial reviewer audits the draft.
|
|
51
|
+
|
|
52
|
+
## Current backend support
|
|
53
|
+
|
|
54
|
+
ORA 0.1.0 currently supports one LLM backend:
|
|
55
|
+
|
|
56
|
+
- **LLM backend:** DeepSeek API
|
|
57
|
+
- **Search and scraping backend:** Firecrawl
|
|
58
|
+
|
|
59
|
+
The default model names are `deepseek-v4-flash` for research and writing, and `deepseek-v4-pro` for planning and review. Other LLM backends are not currently supported.
|
|
60
|
+
|
|
61
|
+
## Installation
|
|
62
|
+
|
|
63
|
+
Install from PyPI:
|
|
64
|
+
|
|
65
|
+
```bash
|
|
66
|
+
pip install open-research-agent
|
|
67
|
+
```
|
|
68
|
+
|
|
69
|
+
The primary CLI command is `open-research-agent`. The shorter `ora` command is also installed as a convenience alias.
|
|
70
|
+
|
|
71
|
+
Install from source for development:
|
|
72
|
+
|
|
73
|
+
```bash
|
|
74
|
+
git clone https://github.com/cameronmpalmer/open-research-agent.git
|
|
75
|
+
cd open-research-agent
|
|
76
|
+
python3 -m venv .venv
|
|
77
|
+
source .venv/bin/activate
|
|
78
|
+
pip install -e ".[dev]"
|
|
79
|
+
```
|
|
80
|
+
|
|
81
|
+
## Configuration
|
|
82
|
+
|
|
83
|
+
Set the required API keys:
|
|
84
|
+
|
|
85
|
+
```bash
|
|
86
|
+
export DEEPSEEK_API_KEY="your-deepseek-api-key"
|
|
87
|
+
export FIRECRAWL_API_KEY="your-firecrawl-api-key"
|
|
88
|
+
```
|
|
89
|
+
|
|
90
|
+
Create a default config file:
|
|
91
|
+
|
|
92
|
+
```bash
|
|
93
|
+
open-research-agent config --init
|
|
94
|
+
```
|
|
95
|
+
|
|
96
|
+
Show the active configuration and intensity levels:
|
|
97
|
+
|
|
98
|
+
```bash
|
|
99
|
+
open-research-agent config --show
|
|
100
|
+
```
|
|
101
|
+
|
|
102
|
+
The config file is stored at:
|
|
103
|
+
|
|
104
|
+
```text
|
|
105
|
+
~/.ora/config.yaml
|
|
106
|
+
```
|
|
107
|
+
|
|
108
|
+
## Quick start
|
|
109
|
+
|
|
110
|
+
Preview a research plan without running the full pipeline:
|
|
111
|
+
|
|
112
|
+
```bash
|
|
113
|
+
open-research-agent plan "What are the tradeoffs between Rust and Go for backend services?"
|
|
114
|
+
```
|
|
115
|
+
|
|
116
|
+
Run a standard research task:
|
|
117
|
+
|
|
118
|
+
```bash
|
|
119
|
+
open-research-agent research "What are the tradeoffs between Rust and Go for backend services?" --intensity 2
|
|
120
|
+
```
|
|
121
|
+
|
|
122
|
+
Run deeper research with adversarial review:
|
|
123
|
+
|
|
124
|
+
```bash
|
|
125
|
+
open-research-agent research "What are the tradeoffs between Rust and Go for backend services?" --intensity 4
|
|
126
|
+
```
|
|
127
|
+
|
|
128
|
+
Save to an explicit file:
|
|
129
|
+
|
|
130
|
+
```bash
|
|
131
|
+
open-research-agent research "AI memory systems" --output ai-memory-systems.md
|
|
132
|
+
```
|
|
133
|
+
|
|
134
|
+
Print only to stdout and do not save a report file:
|
|
135
|
+
|
|
136
|
+
```bash
|
|
137
|
+
open-research-agent research "AI memory systems" --no-save
|
|
138
|
+
```
|
|
139
|
+
|
|
140
|
+
## Intensity levels
|
|
141
|
+
|
|
142
|
+
ORA supports five research intensity levels:
|
|
143
|
+
|
|
144
|
+
| Level | Label | Minimum sources | Max rounds | Reviewer |
|
|
145
|
+
|---|---|---:|---:|---|
|
|
146
|
+
| 1 | Quick | 3 | 1 | No |
|
|
147
|
+
| 2 | Standard | 8 | 1 | No |
|
|
148
|
+
| 3 | Thorough | 15 | 2 | No |
|
|
149
|
+
| 4 | Deep | 50 | 3 | Yes |
|
|
150
|
+
| 5 | Exhaustive | 100 | 4 | Yes |
|
|
151
|
+
|
|
152
|
+
Levels 4 and 5 use the adversarial reviewer by default.
|
|
153
|
+
|
|
154
|
+
## Output files
|
|
155
|
+
|
|
156
|
+
By default, `open-research-agent research` saves a timestamped markdown report in the current directory. Generated research reports are local outputs and should not be committed to the repository.
|
|
157
|
+
|
|
158
|
+
Use `--output` to choose a specific path, or `--no-save` to print the report without writing a file.
|
|
159
|
+
|
|
160
|
+
## Development
|
|
161
|
+
|
|
162
|
+
Install development dependencies:
|
|
163
|
+
|
|
164
|
+
```bash
|
|
165
|
+
pip install -e ".[dev]"
|
|
166
|
+
```
|
|
167
|
+
|
|
168
|
+
Run tests:
|
|
169
|
+
|
|
170
|
+
```bash
|
|
171
|
+
pytest
|
|
172
|
+
```
|
|
173
|
+
|
|
174
|
+
Run the CLI locally:
|
|
175
|
+
|
|
176
|
+
```bash
|
|
177
|
+
open-research-agent --help
|
|
178
|
+
ora --help
|
|
179
|
+
python -m ora --help
|
|
180
|
+
```
|
|
181
|
+
|
|
182
|
+
See `CONTRIBUTING.md` for contributor setup and repository hygiene expectations.
|
|
183
|
+
|
|
184
|
+
## License
|
|
185
|
+
|
|
186
|
+
MIT License. See `LICENSE`.
|
|
@@ -0,0 +1,150 @@
|
|
|
1
|
+
# Open Research Agent (ORA)
|
|
2
|
+
|
|
3
|
+
Open Research Agent (ORA) is an open-source multi-agent research CLI. ORA plans research, searches and scrapes web sources, synthesizes findings, and optionally uses an adversarial reviewer for higher-intensity research.
|
|
4
|
+
|
|
5
|
+
Current release: **0.1.0**
|
|
6
|
+
|
|
7
|
+
## What ORA does
|
|
8
|
+
|
|
9
|
+
ORA turns a research question into a sourced markdown report:
|
|
10
|
+
|
|
11
|
+
1. A supervisor drafts a research plan.
|
|
12
|
+
2. The researcher searches and scrapes web sources.
|
|
13
|
+
3. The writer synthesizes findings into a report.
|
|
14
|
+
4. For intensity levels 4 and 5, an adversarial reviewer audits the draft.
|
|
15
|
+
|
|
16
|
+
## Current backend support
|
|
17
|
+
|
|
18
|
+
ORA 0.1.0 currently supports one LLM backend:
|
|
19
|
+
|
|
20
|
+
- **LLM backend:** DeepSeek API
|
|
21
|
+
- **Search and scraping backend:** Firecrawl
|
|
22
|
+
|
|
23
|
+
The default model names are `deepseek-v4-flash` for research and writing, and `deepseek-v4-pro` for planning and review. Other LLM backends are not currently supported.
|
|
24
|
+
|
|
25
|
+
## Installation
|
|
26
|
+
|
|
27
|
+
Install from PyPI:
|
|
28
|
+
|
|
29
|
+
```bash
|
|
30
|
+
pip install open-research-agent
|
|
31
|
+
```
|
|
32
|
+
|
|
33
|
+
The primary CLI command is `open-research-agent`. The shorter `ora` command is also installed as a convenience alias.
|
|
34
|
+
|
|
35
|
+
Install from source for development:
|
|
36
|
+
|
|
37
|
+
```bash
|
|
38
|
+
git clone https://github.com/cameronmpalmer/open-research-agent.git
|
|
39
|
+
cd open-research-agent
|
|
40
|
+
python3 -m venv .venv
|
|
41
|
+
source .venv/bin/activate
|
|
42
|
+
pip install -e ".[dev]"
|
|
43
|
+
```
|
|
44
|
+
|
|
45
|
+
## Configuration
|
|
46
|
+
|
|
47
|
+
Set the required API keys:
|
|
48
|
+
|
|
49
|
+
```bash
|
|
50
|
+
export DEEPSEEK_API_KEY="your-deepseek-api-key"
|
|
51
|
+
export FIRECRAWL_API_KEY="your-firecrawl-api-key"
|
|
52
|
+
```
|
|
53
|
+
|
|
54
|
+
Create a default config file:
|
|
55
|
+
|
|
56
|
+
```bash
|
|
57
|
+
open-research-agent config --init
|
|
58
|
+
```
|
|
59
|
+
|
|
60
|
+
Show the active configuration and intensity levels:
|
|
61
|
+
|
|
62
|
+
```bash
|
|
63
|
+
open-research-agent config --show
|
|
64
|
+
```
|
|
65
|
+
|
|
66
|
+
The config file is stored at:
|
|
67
|
+
|
|
68
|
+
```text
|
|
69
|
+
~/.ora/config.yaml
|
|
70
|
+
```
|
|
71
|
+
|
|
72
|
+
## Quick start
|
|
73
|
+
|
|
74
|
+
Preview a research plan without running the full pipeline:
|
|
75
|
+
|
|
76
|
+
```bash
|
|
77
|
+
open-research-agent plan "What are the tradeoffs between Rust and Go for backend services?"
|
|
78
|
+
```
|
|
79
|
+
|
|
80
|
+
Run a standard research task:
|
|
81
|
+
|
|
82
|
+
```bash
|
|
83
|
+
open-research-agent research "What are the tradeoffs between Rust and Go for backend services?" --intensity 2
|
|
84
|
+
```
|
|
85
|
+
|
|
86
|
+
Run deeper research with adversarial review:
|
|
87
|
+
|
|
88
|
+
```bash
|
|
89
|
+
open-research-agent research "What are the tradeoffs between Rust and Go for backend services?" --intensity 4
|
|
90
|
+
```
|
|
91
|
+
|
|
92
|
+
Save to an explicit file:
|
|
93
|
+
|
|
94
|
+
```bash
|
|
95
|
+
open-research-agent research "AI memory systems" --output ai-memory-systems.md
|
|
96
|
+
```
|
|
97
|
+
|
|
98
|
+
Print only to stdout and do not save a report file:
|
|
99
|
+
|
|
100
|
+
```bash
|
|
101
|
+
open-research-agent research "AI memory systems" --no-save
|
|
102
|
+
```
|
|
103
|
+
|
|
104
|
+
## Intensity levels
|
|
105
|
+
|
|
106
|
+
ORA supports five research intensity levels:
|
|
107
|
+
|
|
108
|
+
| Level | Label | Minimum sources | Max rounds | Reviewer |
|
|
109
|
+
|---|---|---:|---:|---|
|
|
110
|
+
| 1 | Quick | 3 | 1 | No |
|
|
111
|
+
| 2 | Standard | 8 | 1 | No |
|
|
112
|
+
| 3 | Thorough | 15 | 2 | No |
|
|
113
|
+
| 4 | Deep | 50 | 3 | Yes |
|
|
114
|
+
| 5 | Exhaustive | 100 | 4 | Yes |
|
|
115
|
+
|
|
116
|
+
Levels 4 and 5 use the adversarial reviewer by default.
|
|
117
|
+
|
|
118
|
+
## Output files
|
|
119
|
+
|
|
120
|
+
By default, `open-research-agent research` saves a timestamped markdown report in the current directory. Generated research reports are local outputs and should not be committed to the repository.
|
|
121
|
+
|
|
122
|
+
Use `--output` to choose a specific path, or `--no-save` to print the report without writing a file.
|
|
123
|
+
|
|
124
|
+
## Development
|
|
125
|
+
|
|
126
|
+
Install development dependencies:
|
|
127
|
+
|
|
128
|
+
```bash
|
|
129
|
+
pip install -e ".[dev]"
|
|
130
|
+
```
|
|
131
|
+
|
|
132
|
+
Run tests:
|
|
133
|
+
|
|
134
|
+
```bash
|
|
135
|
+
pytest
|
|
136
|
+
```
|
|
137
|
+
|
|
138
|
+
Run the CLI locally:
|
|
139
|
+
|
|
140
|
+
```bash
|
|
141
|
+
open-research-agent --help
|
|
142
|
+
ora --help
|
|
143
|
+
python -m ora --help
|
|
144
|
+
```
|
|
145
|
+
|
|
146
|
+
See `CONTRIBUTING.md` for contributor setup and repository hygiene expectations.
|
|
147
|
+
|
|
148
|
+
## License
|
|
149
|
+
|
|
150
|
+
MIT License. See `LICENSE`.
|
|
File without changes
|