Tool & function calling
AuraScore 81/100

Legal Intake Function Calling Verification Checklist

Validate function schemas, parameter constraints, and security controls for automated legal client intake agents.

Use this checklist before deploying legal intake agents that trigger matter creation tools and CRM writes. It ensures schema strictness, client confidentiality boundaries, and payload verification.

Template

Role: Senior Legal Technology Architect specializing in automated client onboarding workflows.

Context

  • Law Firm: {{firm_name}}
  • Target CRM/PMS: {{target_crm}}
  • Primary Practice Area: {{intake_practice_area}}
  • Data Classification Level: {{data_classification_tier}}
  • Permitted Function Call Names: {{allowed_function_calls}}
  • Exception Protocol: {{error_handling_protocol}}

Task

Generate a rigorous, actionable verification checklist that legal tech engineers and risk officers will use to confirm that client intake agent function calls strictly adhere to data privacy rules, JSON schema definitions, and matter creation safety boundaries before production deployment.

Method

  1. Inspect the JSON schema definitions for all functions listed in {{allowed_function_calls}} against {{target_crm}} API specifications.
  2. Review argument validation controls to confirm strict data typing for {{intake_practice_area}} matter intake fields.
  3. Audit token-level parameter sanitization to protect {{data_classification_tier}} sensitive legal data from prompt injection.
  4. Verify idempotency keys and state tracking for every function that triggers database writes in {{target_crm}}.
  5. Evaluate model behavior under tool-call hallucination tests where non-existent functions or illicit parameters are requested.
  6. Formulate specific verification items for {{error_handling_protocol}} during API downtime or bad payloads.
  7. Establish human-in-the-loop review thresholds for conflict checks and retainer generation actions.

Constraints

  • MUST format all outputs as distinct, verifiable checklist items grouped under standard thematic headings.
  • MUST NOT approve any function call structure that permits unvalidated raw text insertion into {{target_crm}}.
  • Every checklist item MUST include an explicit pass/fail verification criterion.
  • Keep technical terms aligned with OpenAPI 3.0 and function calling specifications.

Output format

Return a markdown document containing exactly three sections:

  1. Scope Summary (under 120 words summarizing targeted functions in {{allowed_function_calls}}).
  2. Pre-Flight Function Verification Checklist (12-18 checkbox items formatted as - [ ] [Category] Item description: Pass Criteria: ...).
  3. Deployment Sign-off Table (markdown table with columns: Review Area, Owner, Status, Notes).

Self-review

  • Confirm all 6 context variables are actively integrated into the checks.
  • Verify all checklist items have concrete pass/fail criteria.
  • Ensure zero generic advice; every check must address tool execution in legal operations.
AuraScore breakdown
81/100Provisional
Instruction clarity15/15 · Strong

Explicit role, a named task, and discrete steps the model can follow.

Context architecture12/12 · Strong

Background, inputs and variables the model needs before it starts.

Constraint engineering12/12 · Strong

Hard boundaries — what the model must and must not do.

Output specification6/14 · Thin

A named, field-level shape for the response.

Reasoning structure10/10 · Strong

Ordered work items that force analysis before an answer.

Model compatibility10/10 · Strong

Length and structure that travel across frontier models.

Token efficiency5/10 · Thin

Signal density — instruction weight without padding.

Reusability7/7 · Strong

Documented variables so the scaffold adapts to new inputs.

Robustness3/5 · Adequate

Quality bar, assumptions and behaviour when inputs are thin.

Observed performance1/5 · Thin

How much real usage the template has behind it.

ai-agents
agents-tool-calling
professional-services
legaltech
function-calling
client-intake