Spanda

Mesh Security

Mesh communication reuses Spanda secure messaging — mesh does not bypass signing, encryption policy, or trust boundaries.

Message envelope

Every MeshMessage includes: message_id, source_entity, target_entity, target_capability, route, priority, ttl, timestamp, nonce, signature, encryption_required, trust_requirement, payload_hash, and optional SignedMessage envelope.

Validation

The mesh security layer rejects:

• Unsigned high-risk commands
• Untrusted relays for safety-critical routes
• Replayed messages (nonce registry)
• Expired messages (TTL)
• Routes through compromised or low-trust entities

Trust requirements

MeshTrustRequirement controls minimum_trust_score, require_signed, require_identity_match, and block_untrusted_relays.

Regression tests

See crates/spanda-entity-mesh/tests/security_regression.rs.

This site is open source. Improve this page.