Skip to main content
Lab Grimoire
TW EN
☕ Coffee
Spellbook / OpenPlanter ↗ Visit project
ai-tools

OpenPlanter

An open-source recursive LLM-powered investigation agent that performs entity resolution across heterogeneous datasets (corporate filings, campaign finance, government contracts) and renders knowledge graphs in real-time via Cytoscape.js. Built with Tauri 2, MIT licensed, positioned as Palantir's community edition. 1,600 Stars in two months.

Best For

OSINT researchers, investigative journalists, and compliance analysts — anyone who needs to rapidly build entity relationship graphs from multiple public datasets. Particularly suited for individuals or small teams who need Palantir-grade entity resolution capabilities but cannot afford enterprise licensing.

How I Actually Use It

Studied its three-panel desktop layout (Session sidebar / Chat pane / Knowledge Graph) implementation, and how Cytoscape.js integrates with Tauri IPC for real-time graph updates. The recursive agent depth visualization (Depth 0 → Depth 4) is also a noteworthy UX pattern worth referencing.

Where It Is Strong

  • Main agent spawns sub-agents, which can delegate further (max depth 4), breaking through single context window limits for complex investigations
  • Cytoscape.js real-time knowledge graph: Color-coded entity nodes, three layouts (force-directed/hierarchical/circular), live updates with solid visual quality
  • Tauri 2 cross-platform desktop app with pre-built installers for macOS/Windows/Linux
  • Supports OpenAI / Anthropic / OpenRouter / Cerebras / Ollama (local), not locked to any provider
  • Automatically generates Markdown documentation during investigation (Wiki curator drawer). Results are traceable and shareable
  • MIT license, fully open, commercially usable

Where It Fails

  • Agent has full terminal execution capability. Running without isolation allows arbitrary commands on host, so Docker deployment is mandatory
  • v0.1.x is designed as a personal tool with no user login or permissions
  • LLM API calls send data fragments to third-party services. Use local Ollama for sensitive investigations
  • Early version: v0.1.1, API may have breaking changes
  • Not suitable for: enterprise environments handling sensitive data, teams requiring multi-user collaboration

Pricing, Difficulty, and Risk

  • Completely free (MIT). LLM API costs vary by usage (free Ollama option available)
  • Download the pre-built binary and start. Configure an API key to begin investigating. Secure deployment requires Docker understanding
  • Risk: Shell execution security (Docker isolation required), early version instability, third-party LLM data exposure

Verdict

OpenPlanter replaces manual cross-referencing with AI agents and breaks single-model limitations through recursive delegation. The real-time knowledge graph visualization is well executed. The biggest caveat is security: the agent's shell execution permissions mean Docker deployment is mandatory, and v0.1.x stability still needs to prove itself. Good for personal research and exploration. Enterprise environments should wait for the security framework to mature.

Source