agentforce-probe

v0.3.0 suspicious
7.0
High Risk

Run automated tests against Salesforce Agentforce agents (External + Internal Copilot) and score them into evidence — fully local, privacy-first, no API key required.

🤖 AI Analysis

Final verdict: SUSPICIOUS

The package shows unusual behavior with high shell execution risk and moderate network interaction, suggesting potential malicious activities. Despite no direct evidence of obfuscation, credential theft, or extensive metadata issues, the overall pattern is concerning.

  • High shell execution risk
  • Moderate network interaction
Per-check LLM notes
  • Network: The network call patterns suggest the package may be designed to make HTTP requests, which could be normal if it's intended to probe or interact with external services.
  • Shell: The shell execution patterns indicate that the package can execute commands on the host system, which is highly unusual and potentially indicative of malicious intent.
  • Obfuscation: No obfuscation patterns detected, indicating low risk of malicious intent.
  • Credentials: No credential harvesting patterns detected, indicating secure handling of sensitive information.
  • Metadata: The repository is new with no activity indicators, suggesting potential unreliability.

🔬 Heuristic Checks

⚠ Outbound Network Calls score 7.5

Found 5 network call pattern(s)

  • range(retries): req = urllib.request.Request(url, data=body_bytes, method=method) for k,
  • try: with urllib.request.urlopen(req, timeout=timeout) as resp: retur
  • pt in range(3): req = urllib.request.Request(url, data=form, method="POST") req.add_heade
  • try: with urllib.request.urlopen(req, timeout=timeout) as resp: paylo
  • range(retries): req = urllib.request.Request(url, data=body, method="POST") for k, v in h
✓ Code Obfuscation

No obfuscation patterns detected

⚠ Shell / Subprocess Execution score 2.0

Found 1 shell execution pattern(s)

  • "sf"] + list(args) proc = subprocess.run( cmd, capture_output=True, text=True
✓ 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 score 5.0

Git history flags: Repository created very recently: 4 day(s) ago (2026-06-01T23:44:06Z)

  • Repository created very recently: 4 day(s) ago (2026-06-01T23:44:06Z)
  • Repository has zero stars and zero forks
⚠ Maintainer History score 2.0

1 maintainer concern(s) found

  • Author "Ray Kuo" appears to have only 1 package on PyPI (new or inactive account)
✓ Known CVE Vulnerabilities

No known vulnerabilities found in OSV database.

💡 AI App Starter Prompt

Use this prompt to build a project with agentforce-probe 
Develop a Python-based mini-application named 'AgentForceTester' that leverages the 'agentforce-probe' package to automate the testing of Salesforce Agentforce agents, both External and Internal Copilots, in a privacy-first manner without requiring any API keys. The application should perform the following steps and include the specified features:

1. **Setup**: Initialize the application by installing the 'agentforce-probe' package. Ensure that the setup process includes instructions on how to configure the environment for running the tests.
2. **Test Scenario Creation**: Allow users to define test scenarios that include different types of queries or tasks they want to test against the Agentforce agents. These scenarios should cover various aspects such as knowledge retrieval, task execution, and conversational capabilities.
3. **Execution**: Implement a feature to run these test scenarios against the Agentforce agents locally. The application should support both batch and single-run test executions.
4. **Scoring and Feedback**: After executing each test scenario, use the 'agentforce-probe' scoring mechanism to evaluate the performance of the agents. Provide detailed feedback on each test result, highlighting strengths and weaknesses.
5. **Report Generation**: Create a report summarizing all the test results, including overall scores and individual scenario outcomes. This report should be easily readable and exportable in formats like PDF or CSV.
6. **User Interface**: Develop a simple command-line interface (CLI) for interacting with the application. Additionally, consider adding a basic web interface using Flask or Django for more interactive user experience.
7. **Privacy Assurance**: Emphasize the privacy-first approach of the application by ensuring that all data processing occurs locally and no personal or sensitive information is transmitted or stored externally.
8. **Documentation**: Provide comprehensive documentation detailing how to install, configure, and use the application effectively. Include examples and best practices for creating effective test scenarios.

The 'agentforce-probe' package is utilized throughout the application to handle the automation of test runs, scoring of agent responses, and generation of evidence-based feedback. Ensure that the integration with 'agentforce-probe' is seamless and that the application takes full advantage of its capabilities to provide accurate and insightful evaluations of Agentforce agents.