workX Coding Agent
Enterprise terminal coding agent: CLI / IDE integration and on-premises delivery, bringing AI coding productivity inside your security boundary.
Request Trial →Why workX Coding Agent
Pain point: Public-cloud AI coding tools fail code security and compliance requirements
Full on-premises delivery — code and conversations never leave your boundary
Pain point: AI tool usage and cost are invisible and uncontrollable
Team-level metering, quotas, and cost attribution via Cangyun AI Gateway
Pain point: Open-source agents evolve fast; in-house forks are hard to maintain
Commercial distribution tracks upstream; upgrades and customization handled by Cangyun
Pain point: Team coding standards are hard to enforce consistently
Built-in standards injection — agent output follows your coding conventions
Features
Terminal-first Agent Experience
Complete chat–code–run–verify loop in the terminal, with long-running autonomous tasks.
IDE & Enterprise Integration
app-server protocol connects IDE plugins and internal platforms — one core, many clients.
Enterprise Permissions & Security
Command approval, sandbox policies, and repository config governance.
Bring Your Own Models
Connect self-hosted model services or multiple vendors routed through the AI Gateway.
License Management
Offline Ed25519 local-verification licensing for air-gapped environments.
Continuous Upstream Sync
Automated upstream drift detection and upgrade gates keep pace with the community.
Deployment & Delivery
- ✓On-premises deployment, including fully air-gapped environments
- ✓CLI distributed via private npm registry with unified version governance
- ✓Bundled with Cangyun AI Gateway for out-of-the-box metering and control
- ✓Consulting and tooling for enterprise coding standards (AGENTS.md)
FAQ
Which models does workX support?
Any OpenAI-protocol-compatible model service; multiple vendors can be unified and routed via Cangyun AI Gateway.
Does code ever leave our network?
With on-premises deployment, code and conversations stay entirely within your boundary — no third-party services involved.
How do we control team usage cost?
Cangyun AI Gateway provides multi-tenant metering and rate limiting, with per-team quotas and billing.