v0.4.0GitHub

Architecture & security

Threat Model

Blekline ingress, sidecar, and extension threat model for security review.

View as Markdown

This page summarizes how Blekline addresses common AI agent security threats at the execution boundary (Layer 4).

Assets

  • Workspace policy, audit events, API keys (blw_*)
  • Prompt and tool-call content transiting ingress/sidecar
  • Trust Vault ciphertext (NHIM sidecar, in-VPC)

Trust boundaries

See Trust Boundaries. Summary:

  • Outside perimeter: model providers, untrusted web content, user prompts
  • Inside perimeter: Blekline control plane, sidecar, your VPC tool endpoints

Threats & mitigations

ThreatMitigation
Secret leakage to LLMLocal mask + cloud Azure PII; Trust Vault tokenization (NHIM)
Prompt injection → destructive toolsLineage Firewall + YAML/runtime policy; simulator
Over-privileged tool callsMCP tool policy, enforceToolCallLocally, sidecar enforce
Shadow AI agentsIngress/sidecar at agent boundary; audit events
Downstream MCP impersonationMCP identity pinning

Browser extension surface

Enterprise deployments should treat browser extensions as high-trust surfaces. Compensating controls: workspace policy, optional MDM distribution, least-privilege workspace tokens.

Out of scope (current)

  • Full CASB / network DLP replacement
  • SOC 2 Type II attestation (roadmap)
  • Cryptographic SPIFFE proof in OSS (roadmap — see NHIM)