Update block docs for: hubspot/engagement.md

This commit is contained in:
Bently
2025-01-13 10:06:18 +00:00
parent 43ca817a85
commit 2c9baa6966

View File

@@ -0,0 +1,32 @@
<file_name>autogpt_platform/backend/backend/blocks/jina/fact_checker.md</file_name>
## Fact Checker
### What it is
A specialized tool that evaluates the truthfulness of statements using Jina AI's Grounding API technology.
### What it does
This block analyzes a given statement and determines its factuality, providing a detailed assessment that includes both a numerical score and a reasoned explanation for its conclusion.
### How it works
The block takes a statement and processes it through Jina AI's fact-checking system. It securely sends the statement to Jina's API, which analyzes the content and returns a comprehensive evaluation. The system then processes this evaluation to provide clear, actionable results about the statement's truthfulness.
### Inputs
- Statement: The text statement that needs to be verified for factuality
- Credentials: Secure authentication details required to access the Jina AI service
### Outputs
- Factuality: A numerical score indicating how factual the statement is
- Result: A simple true/false determination of the statement's factuality
- Reason: A detailed explanation of why the statement was classified as true or false
- Error: Any error messages that might occur during the fact-checking process
### Possible use case
A content moderation team could use this block to verify claims made in user-generated content on a social media platform. For example, if a user posts a statement about a historical event, the team could quickly check its accuracy before the content goes live. This helps maintain information integrity and combat the spread of misinformation.
### Additional Notes
- The block is categorized under the SEARCH category, indicating its role in information verification and retrieval
- It utilizes secure API authentication to ensure reliable and authorized access to the fact-checking service
- The system provides both quantitative (factuality score) and qualitative (reason) assessments, offering a comprehensive evaluation of each statement