Wealth Management Agent Tool Authorization Matrix

Role: Senior FinTech Security Architect specializing in autonomous agent governance and API security.

Context

• Financial Institution: {{institution_name}}
• Wealth Advisory Tier: {{client_tier}}
• Core Banking API: {{core_banking_engine}}
• Regulatory Framework: {{compliance_standard}}
• Transaction Threshold: {{transaction_threshold}}
• Authentication Protocol: {{authentication_level}}

Task

Generate a comprehensive tool authorization and risk gating matrix that defines exact permission levels, execution modes, payload parameters, and circuit breakers for AI agent function calls across customer wealth accounts.

Method

1. Enumerate all core function calls required by the agent across the {{core_banking_engine}} interface.
2. Classify each function into read-only queries, low-risk adjustments, or high-risk capital movements against {{transaction_threshold}}.
3. Determine the required execution mode (autonomous execution, user-confirmed prompt, or dual-custody human sign-off) tailored to {{client_tier}}.
4. Map each function to specific authorization scopes mandated under {{compliance_standard}}.
5. Define validation constraints for inbound function arguments and schema types.
6. Specify fallback routines and circuit breakers for failed authentication or schema mismatch under {{authentication_level}}.
7. Format the complete policy into a structured evaluation matrix.

Constraints

• MUST enforce explicit human confirmation on any state-changing function call exceeding {{transaction_threshold}}.
• MUST NOT permit plain-text account identifiers or unmasked personally identifiable information in tool arguments.
• Restrict all execution scopes to the minimum viable permissions required for {{client_tier}}.
• All matrix entries must map directly to {{compliance_standard}} compliance controls.

Output format

1. Architectural Scope

A 2-paragraph overview detailing the agent interface and permission boundaries.

2. Tool Authorization Matrix

A Markdown table containing the following columns: | Function Name | Description | Invocation Mode | Required Auth Scope | Param Constraints | Failure / Circuit Breaker | (Include 6 to 8 granular tool entries covering balance checks, allocations, rebalancing, fund transfers, and report exports).

3. Execution Guardrails

A bulleted list of 4 policy rules governing edge-case session termination.

Self-review

• Verify all 6 matrix columns are populated with specific technical arguments rather than generic placeholders.
• Confirm every function call exceeding {{transaction_threshold}} requires human sign-off.
• Ensure strict compliance with {{compliance_standard}} requirements.