vexjoy-agent
VexJoy AI Agent with Jev Intelligent Routing - /do routes plain-English requests to the right specialist agent and gates the work with reviews, tests, and a lea
What is vexjoy-agent?
VexJoy AI Agent with Jev Intelligent Routing - /do routes plain-English requests to the right specialist agent and gates the work with reviews, tests, and a learning loop.
Overview of vexjoy-agent
What it is
VexJoy Agent is an AI agent toolkit that connects plain-English requests to specialist agents, skills, and workflows. It uses /do to route requests to the appropriate knowledge and tools, with hooks enforcing checks and scripts handling repeatable work. The toolkit includes 43 agents, 62 skills, 79 hooks, and 156 scripts. It works across Claude Code (/do), Codex ($do), Factory (/do), and Reasonix (/do).
Key features
/doroutes plain-English requests to the right specialist agent, skill, and pipeline.- 43 agents with domain knowledge, 62 reusable skills, 79 lifecycle hooks, and 156 scripts.
/d(Jev-powered router) classifies requests, checks intent preservation, and dispatches with a Jev receipt.- Anti-rationalization gates require evidence (test output, exit codes) rather than confidence.
- Content engine researches, drafts, checks 397 writing patterns, and adapts content for platforms;
/htmlproduces self-contained reports, decks, prototypes, charts, or diagrams. - Multi-runtime support: Claude Code, Codex, Factory, Reasonix, with parity adapters for hooks and commands.
- Automated nightly evolution loop (
/evolve) ran regularly through mid-May 2026; currently dormant.
Use cases
- Debugging and fixing code (e.g., Go test failures, flaky tests).
- Automated code review and PR creation with verification gates.
- Knowledge work: research, writing, and platform adaptation.
- Multi-runtime agent orchestration across different CLI tools.
- Intent-preserving task routing with Jev alignment checks.
Getting started
bash git clone https://github.com/notque/vexjoy-agent.git ~/vexjoy-agent cd ~/vexjoy-agent ./install.sh For partial installation, run ./install.sh --configure or edit .local/profile.yaml. See docs/start-here.md for full setup.
License and pricing
Licensed under MIT. Open source.
Based on the project's public README and repository metadata.