Private beta · Request a scoped Agent Firewall evaluation

Trust Center

Security claims you can verify.

RaksHex is built to give teams control over AI access without turning customer prompts and provider keys into another source of risk.

Credential boundaries

Workspace credentials are encrypted before storage. List APIs return masked metadata and fingerprints, not encrypted values or plaintext keys. Revocation actions are audited.

Prompt-minimizing design

Discovery ingests masked findings. Gateway policy evaluation can redact PII and records audit metadata without retaining raw prompts by default.

Customer control

Workspace data is scoped by membership and can be exported or deleted through the product and support process. Retention and private-data-plane requirements are agreed for enterprise deployments.

Evidence, not badges

RaksHex maps product controls to common frameworks and produces evidence. We do not claim a certification or independent audit until that assessment is complete and published.

What a security review can inspect

  • Workspace authorization checks and tenant-isolation test coverage.
  • Credential encryption, fingerprinting, lifecycle state, and sensitive-action audit trails.
  • Prompt-retention defaults, redaction behavior, provider capability limits, and data provenance labels.
  • Deployment architecture for SaaS, private relay, and self-hosted enterprise environments.

For a DPA, security questionnaire, architecture review, or private deployment requirements, email privacy@rakshex.in, security@rakshex.in, or legal@rakshex.in before connecting production accounts.