aquin

v0.0.5 suspicious
6.0
Medium Risk

Record training runs locally and push to Aquin for post-hoc inspection — loss curves, SAE diffs, model diffs, and more. Browse and download public SAEs via aquin list saes / aquin pull sae.

🤖 AI Analysis

Final verdict: SUSPICIOUS

The package exhibits moderate risk due to its network activity and potential for unauthorized data transmission. Additionally, low maintainer engagement and poor metadata quality raise concerns about transparency and intent.

  • Moderate network risk
  • Low maintainer engagement
  • Poor metadata quality
Per-check LLM notes
  • Network: The package makes several network calls that could be used for unauthorized data transmission or C2 communication.
  • Shell: No shell execution patterns were detected.
  • Obfuscation: No obfuscation patterns detected, indicating low risk.
  • Credentials: No credential harvesting patterns detected, indicating low risk.
  • Metadata: The package shows signs of low maintainer engagement and poor metadata quality, which may indicate a lack of transparency or intent.

📦 Package Quality Overall: Low (2.8/10)

○ Low Test Suite 1.0

No test suite detected

  • No test files or test-runner configuration detected
◈ Medium Documentation 5.0

Some documentation present

  • Detailed PyPI description (3899 chars)
○ Low Contributing Guide 2.0

No contributing guide or governance files found

  • No CONTRIBUTING, CODE_OF_CONDUCT, or governance files found
◈ Medium Type Annotations 5.0

Partial type annotation coverage

  • 27 type-annotated function signatures detected in source
○ Low Multiple Contributors 1.0

Unable to verify contributor count: no GitHub repository found

  • No GitHub repository linked — contributor count unavailable

🔬 Heuristic Checks

⚠ Outbound Network Calls score 9.0

Found 6 network call pattern(s)

  • no body) auth_resp = requests.get( f"{base_url}/api/sdk/push", params=
  • rb") as f: resp = requests.post( upload["url"], data=f,
  • ("/") try: resp = requests.get( f"{base_url}/api/sdk/whoami", heade
  • import requests resp = requests.get(f"{_base_url()}/api/public/saes", timeout=15) resp.raise
  • import requests resp = requests.get(f"{_base_url()}/api/public/saes/{id}", timeout=15) if re
  • e, exist_ok=True) resp = requests.get( f"{_base_url()}/api/public/saes/{id}/pull",
✓ Code Obfuscation

No obfuscation patterns detected

✓ Shell / Subprocess Execution

No shell execution patterns detected

✓ Credential Harvesting

No credential harvesting patterns detected

✓ Typosquatting

No typosquatting candidates detected

✓ Registered Email Domain

No author email provided

✓ Suspicious Page Links

All external links appear legitimate

✓ Git Repository History

No GitHub repository linked

  • No GitHub repository link found
⚠ Maintainer History score 6.0

3 maintainer concern(s) found

  • Author name is missing or very short
  • Author "" appears to have only 1 package on PyPI (new or inactive account)
  • Package has no PyPI classifiers (low effort / metadata quality)
✓ Known CVE Vulnerabilities

No known vulnerabilities found in OSV database.

💡 AI App Starter Prompt

Use this prompt to build a project with aquin 
Create a Python-based machine learning experiment tracker and inspector using the 'aquin' package. Your task is to develop a command-line tool that allows users to easily record their machine learning training runs locally and then push these records to the Aquin platform for post-hoc analysis. This tool should also enable users to browse and download publicly available SAEs (Single-Agent Environments) from Aquin. Here’s a step-by-step guide on what your application should achieve:

1. **Setup and Installation**: Ensure the application can be installed via pip and includes dependencies like 'aquin'. Provide clear instructions on setting up the environment.
2. **Local Training Run Recording**: Develop functionality to record details of machine learning training runs locally. This includes capturing metrics such as loss curves, model weights, and other relevant information.
3. **Pushing to Aquin**: Implement a feature to upload the recorded training data to Aquin. Users should be able to specify which parts of the training run they want to upload (e.g., specific epochs, model checkpoints).
4. **Post-Hoc Analysis**: Once uploaded, users should be able to use Aquin's features for post-hoc inspection, such as comparing different models or analyzing loss curves over time.
5. **Browsing and Downloading Public SAEs**: Integrate a command that allows users to list all available public SAEs on Aquin and another command to download selected SAEs.
6. **User Interface**: Design a simple and intuitive command-line interface for interacting with the application.
7. **Documentation and Examples**: Include comprehensive documentation and example usage scenarios to help new users get started quickly.
8. **Testing and Validation**: Ensure thorough testing of all functionalities, including local recording, uploading, and downloading from Aquin.

This project aims to streamline the process of tracking and inspecting machine learning experiments, making it easier for researchers and developers to manage their projects efficiently.

💬 Discussion Feed

Leave a comment

No discussion yet. Be the first to share your thoughts!