OPREP: Open-Source OSINT Analysis Platform Powered by Local LLMs
Jesus_Morty · reddit · 2026-08-05
A developer released OPREP, a self-hosted open-source intelligence (OSINT) analysis platform. The project utilizes locally operated text, vision, and embedding models to assist in multi-source intelligence analysis.
Key Features:
- Builds a persistent entity and evidence graph with semantic search
- Source-weighted correlation into incidents, warnings, and campaigns
- Tracks provenance, confidence, contradictions, and intelligence gaps
- Image geolocation using a local vision model
- Built-in structured analysis methods including ACH, red teaming, and link analysis
- Keeps underlying evidence available to the operator; models only assist with enrichment, correlation, and drafting
The project is in its early v0.1.0 stage, with the author running models locally on a single RTX 3090.
More from coding & agent
- You.com Search API Adds x402 Support for Autonomous AI Agent Payments — kleffew94 · 2026-08-05
- Goodfire AI Launches Silico for Autonomous Long-Horizon Experiments — mathildepapillo · 2026-08-05
- Opinion: RL Environments are all you need for training and evaluating agents — gregd_nlp · 2026-08-05
- llm-checker: inspect model files before loading to prevent crashes from malicious GGUF — tetsuoai · 2026-08-05
- Cloudflare Pay Enables AI Agents to Participate in Online Checkout — davidhoang · 2026-08-05
- Pokee Launches 28B Agentic Model with 10M Token Context, Runs on a Single RTX 4090 — Scobleizer · 2026-08-05