OpenInvest

•
142
•

Orchestrate AI agents in an isolated, auditable committee to make research-grade investment decisions with robust backtesting and transparency.

OpenInvest is a sophisticated, self-hosted investment decision engine designed to empower modern AI agents like Claude Code, Codex, and OpenClaw. It doesn't aim to be a return-amplifying black box but rather an auditing tool that eliminates human cognitive biases and enforces reasoning transparency. By providing a verifiable investment committee, evidence-based reasoning, and long-horizon backtesting, OpenInvest produces auditable decision records. It operates on a philosophy where "your agent knows you, OpenInvest knows investing," offering clean APIs and agent skills that integrate seamlessly without "owning" the user's data or context. The system is rigorously designed to self-falsify its own edge, transparently publishing negative results and focusing on discipline and transparency over directional prediction.

Características Principales

01Agent Skill for seamless integration with Claude Code, Codex, Hermes, and OpenClaw.
0265 GitHub stars
03Multi-Agent Investment Committee with isolated analysis and rebuttal.
04Auditable Decision Trail with clean logs for decision-making.
05Long-Horizon Backtesting with built-in lookahead bias protections.
06Self-Hosted and Zero-Cost execution leveraging local agent resources.

Casos de Uso

01Integrating advanced investment intelligence into personal AI agents.
02Automating daily investment decision reports and portfolio monitoring via GitHub Actions.
03Developing and testing algorithmic trading strategies with research-grade rigor.