OpenFishh

OpenFishh

OpenFishh is a persistent collective intelligence platform that deploys thousands of AI agents to read the open Internet. Unlike chatbots that forget after answering questions, OpenFishh runs a living society of agents around the clock-beliefs are accumulated, information sources are repeatedly evaluated, and contradictory opinions are debated.

17GitHub
Description

##Functional characteristics ###Multi-agent intelligence collection - Support batch generation of intelligence agents with different roles, covering 31 intelligence subject areas - Agents have a division of labor such as 'scout / researcher / cartographer /infilter/ tracker / analyst / qualifier' - Run the daily intelligence cycle in batches by theme and continue to collect information from public RSS feeds ###Belief extraction and knowledge precipitation - Extract belief data in the form of subject, predicate, and object from collected articles - Attach source level, time attribute, confidence interval, and position label to each belief - Automatic entity aggregation, belief storage and basic cross-corroboration ###Research and reporting capabilities - Support the launch of 'Investigate' investigation missions and stream the generation of intelligence reports with a chain of evidence - The report includes a summary of credibility, known unknowns and cited sources for easy review - The generated report will be persisted for subsequent viewing and tracking ###Operational observation capabilities - Provide Society view to view agents, beliefs, entities and contested beliefs - Provide a Cycle page to view real-time progress and statistical results of a complete cycle - Epistemic health that provides Scorecard page assessment belief libraries ###Configure and fill in LLM parameters 1. After opening the application, fill in the relevant LLM information in the configuration wizard 2. If you use OpenRouter, you can keep the default LLM Base URL 3. If you use compatible services such as OpenAI, DeepSeek, and One API, please modify the LLM API Key, LLM Base URL and LLM Model simultaneously. ###Initialize the intelligence society 1. Go to the home page and click `Spawn Your Society` 2. Use the Starter or Medium scale verification process first 3. After initialization is complete, enter the Society page and confirm that agents and beats have been generated ###Run Daily Cycle once 1. Open the Cycle page 2. Click on `Run Cycle Now` 3. Wait for the system to grab RSS, extract beliefs, and complete library writing ###Generating research report 1. Return to the home page and enter the survey questions 2. Click `Investigate` 3. Wait for the system to stream back to the investigation steps, and automatically jump to the report page after completion

Screenshots
Screenshot 1
Screenshot 2
Screenshot 3
Screenshot 4
Screenshot 5
App Information
Version
0.1.0
Package Size
26 KB
Image Size
58.82 MB
Updated
April 27, 2026
Source Code
MohdTalib0
Platform Support
PC
Keywords
AI AgentIntelligence ResearchFastAPIReactSQLiteRSS