**Building the trust and compliance layer for AI agents.
We’re developing open infrastructure that makes agent workflows secure, auditable, and compliant — from redaction to runtime enforcement.
💡 Interested in using Vault Plus in your stack? Reach out at founder@marvisvault.com or Visit Our Website
🔐 Marvis Vault OSS
- Runtime redaction, policy enforcement, audit logging
- CLI + SDK for AI pipelines and agentic apps
- Org Repo →
🧱 Vault Plus (Coming Soon)
- Hosted FastAPI service
- Role-based unmasking and real-time monitoring
- Compliance-grade policy marketplace (GDPR, HIPAA, PCI)
🧪 Experimental Modules
- Tools, internal packages, and R&D projects incubated in
marvis-rnd.
| Name | Description |
|---|---|
marvis-vault |
OSS CLI & SDK for trust enforcement in AI workflows |
vault-cloud (private) |
Vault as a service — deployable runtime agent firewall |
vault-rnd (private) |
Experimental tools, redaction plugins, simulation modules |
We’re actively building and growing the ecosystem:
- 🔧 Start contributing → Open Issues
- 🧪 Run the CLI → Quickstart
- 💬 Share ideas → Open a Discussion
🧠 We believe agentic AI needs a trust kernel — programmable, observable, and secure by design.
Built and led by @abhigyanbiswas, with contributions from trusted collaborators.
License: MIT — open by default.
Vault Plus is commercially supported and in private development.
We’re open to:
- Early adopters and design partners
- Contributors who want to help secure the agentic future
- Startup and enterprise collaborators working on AI infrastructure