Skip to content
Bitcoin Bastion

Security model

Bitcoin Bastion Security

A no-custody, operator-controlled model for serious users and developers. Claims below are scoped to baseline reality and clearly mark future capabilities.

No-Custody Visual Seal

No seed phrases. No private keys. No custodial control. No automatic transaction signing.

Human Confirmation Firewall (demo)

Step 1

Policy trigger

High-risk condition detected.

Step 2

Human review

Operator validates evidence and context.

Step 3

Explicit confirmation

Only then can downstream action proceed.

Demo flow for model explanation; not a claim of fully automated production workflow coverage.

Risk Lens cards

Custody risk lens: kept out-of-scope by design.
Operational risk lens: policy + audit + human gates.
Model risk lens: AI assistance restricted from key control.

Security model diagram

Watch-only dataPolicy EngineHuman Confirmation FirewallExternal signer (future/required)

Bastion remains advisory; signing authority stays external.

Security controls and capability status

No seed phrases

Implemented baseline

Bitcoin Bastion does not accept, process, or request mnemonic seed phrases.

No private key storage

Implemented baseline

Private keys are outside platform scope and must remain in external custody systems.

No custodial control

Implemented baseline

The platform is advisory-only and does not hold funds or signing authority.

No automatic transaction signing

Implemented baseline

No autonomous signing or broadcast pipeline is provided.

Watch-only direction

Implemented baseline

Data surfaces are designed for watch-only intelligence and risk review workflows.

PSBT-first future direction

Planned direction

Future integration direction prioritizes PSBT-compatible boundaries with external signers.

External signing model

Planned direction

Signing should remain external to Bastion via user-controlled hardware/software signers.

Human Confirmation Firewall

Implemented baseline

Risky actions require explicit human confirmation and auditable checkpoints.

Policy Engine

Implemented baseline

Policy constraints drive advisory outputs and operational guardrails.

Audit Log

Implemented baseline

Operator-relevant events are expected to be captured for review and traceability.

Role-based access

Baseline / evolving

Access boundaries are role-oriented and continue maturing with enterprise controls.

AI action restrictions

Implemented baseline

AI-assisted flows are constrained from custody operations and key-material interactions.