Automated Peer Review AI Agent
An end-to-end automation toolkit that ingests research articles (PDF or DOCX), extracts text, performs heuristic quality checks, and produces a full peer-review packet for journal clubs or manuscript review boards.
Key Features
- Multiformat ingestion – reads both PDF and DOCX sources.
- Heuristic appraisal – detects study design, sample size, statistical methods, and flags missing elements (power statements, limitations, precision, etc.).
- Auto-generated collateral:
- Markdown critical analysis report.
- Presentation-ready PPTX deck.
- Structured peer-review DOCX (major/minor comments).
- Annotated text file with inline pseudo track changes.
- Redline-style DOCX with prioritized action items and section-specific rewrite suggestions (works for PDFs by embedding extracted text).
- Batch orchestration – scan a root directory or target a single project folder.
- Extensible CLI – toggle peer-review artifacts, annotations, and redlines independently.
Installation & Distribution Options
-
Local development (recommended)
git clone https://github.com/hssling/Automated_Peer_Review_AI_Agent.git cd Automated_Peer_Review_AI_Agent python -m venv .venv && .venv\Scripts\activate # or source .venv/bin/activate on *nix pip install -e .
Editable installs make iterating on heuristics and outputs straightforward.
-
Pip/Pipx distribution (implemented)
The repository ships with apyproject.toml; runpip install .to build a wheel. For isolated CLI usage,pipx install .creates a self-contained executable environment ideal for automation servers. -
Containerization (suggested)
For fully reproducible deployments (CI agents, on-prem review services), wrap the CLI in a lightweight Python container (e.g., Python 3.11-slim) and mount the articles directory. See “Future Enhancements” for ideas.
CLI Usage
After installation, the CLI peer-review-agent becomes available.
peer-review-agent --root "/path/to/root" [--peer-review] [--annotate] [--redline] [--force]
Common workflows:
- Single project folder
peer-review-agent --folder "D:/Journal club/TB cohort study" --peer-review --annotate --redline
- Batch process every subfolder under a root directory
peer-review-agent --root "/data/articles" --peer-review --annotate
Flags:
--root PATH– scan root subfolders for PDFs/DOCXs.--folder PATH– process a specific folder (overrides--rootdiscovery).--peer-review– create structured DOCX peer-review memo.--annotate– emit annotated text with inline pseudo comments.--redline– generate redline-style DOCX with action plan and rewrite suggestions (works for PDF inputs via extracted text).--force– rebuild outputs even if files already exist.
Output Artifacts
For each source article <stem> the agent produces (based on enabled flags):
<stem>.txt– raw extracted text.<stem>_auto_critical_analysis.md– heuristic report with strengths/gaps.<stem>_auto_appraisal.pptx– slide deck summarizing findings.<stem>_peer_review.docx– major/minor comment log (when--peer-review).<stem>_annotated_comments.txt– inline comments (when--annotate).<stem>_redline_review.docx– prioritized action list + inline suggestions (when--redline).
Derived artifacts are automatically skipped during subsequent runs to avoid recursion.
Requirements
- Python 3.9+
- Dependencies:
PyPDF2,python-docx,python-pptx
Install via pip/venv:
pip install -r requirements.txt
Continuous Integration
A GitHub Actions workflow (.github/workflows/ci.yml) runs linting and unit tests on every push/pull request using Python 3.11. Extend the pipeline with packaging/publishing steps once ready to distribute wheels or Docker images.
Tests
Unit tests live under tests/. Run:
pytest
Roadmap / Future Enhancements
- Publish pre-built wheels to PyPI for
pip install peer-review-agent. - Add optional Docker image for air-gapped deployments.
- Integrate citation parsing and GRADE-style scoring.
- Support configurable templates (custom PPT branding, peer-review rubrics).
Contributions and issue reports are welcome!
Metadata
Release files for peer-review-agent 0.1.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| peer_review_agent-0.1.0.tar.gz | 14.7 kB | Without provenance | Details for peer_review_agent-0.1.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| peer_review_agent-0.1.0-py3-none-any.whl | Python 3 | none | any | Details for peer_review_agent-0.1.0-py3-none-any.whl |
Total release size: 27.6 kB
Release files / peer_review_agent-0.1.0.tar.gz
| Download URL | peer_review_agent-0.1.0.tar.gz |
|---|---|
| Size | 14.7 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
cf73b90a9e2d49f723c15d7a98102c37830c836efff8bab9148d43dcabfac244
|
|
BLAKE2b-256 checksum How to use checksums |
8938097759758bbdb567c14fc40060fbbaac61bb324861e2cfb852b1e2d1aaae
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.14.0
|
Release files / peer_review_agent-0.1.0-py3-none-any.whl
| Download URL | peer_review_agent-0.1.0-py3-none-any.whl |
|---|---|
| Size | 12.9 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
76aee1bb2c581a0ce27c1443414506ddcee96904a1819e6d95242631f6ca2563
|
|
BLAKE2b-256 checksum How to use checksums |
c1421ebda0c7c0a33f45ce4f80d0a244e6af61ef7e84fe57bb57a08f6ba6cdb2
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.14.0
|
Security for peer-review-agent 0.1.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.