Agent instructions
AuraScore 81/100

Autonomous Agent Tool Definition and Orchestration Blueprint

Define robust schema specifications, validation boundaries, and failure-handling strategies for multi-step agent tool calling.

Use this template when building enterprise-grade tool-calling schemas and execution logic for autonomous agents. It creates a complete architectural specification balancing latency, schema safety, and deterministic fallback paths.

Template

Role: Principal Agentic Systems Architect with 12+ years of experience in distributed systems and LLM function-calling architectures.

Context

  • Target Agent: {{agent_name}}
  • Operating Runtime: {{target_environment}}
  • External Integrated APIs: {{external_apis}}
  • Latency Budget: {{latency_budget_ms}}
  • Fault Tolerance Model: {{failure_tolerance_policy}}
  • Regulated / Sensitive Data Boundaries: {{sensitive_data_scope}}

Task

Generate a comprehensive Tool-Calling and Execution Specification Report that defines strict parameter schemas, orchestration rules, stateful context injection, and validation guardrails for {{agent_name}}.

Method

  1. Analyze {{external_apis}} to map discrete function signatures into standardized JSON Schema specifications with explicit property types, enums, and required parameters.
  2. Define pre-invocation parameter validation rules to catch missing, malformed, or out-of-boundary values before invoking runtime calls within {{target_environment}}.
  3. Architect context injection logic specifying how previous tool responses are summarized or filtered to prevent context window bloat.
  4. Formulate deterministic error-handling trees based on {{failure_tolerance_policy}}, detailing automatic retry limits, exponential backoff, and synthetic fallback responses.
  5. Design security and redaction filters that sanitize inputs and outputs against {{sensitive_data_scope}}.
  6. Specify tool selection arbitration criteria to prevent infinite calling loops and enforce execution within {{latency_budget_ms}}.
  7. Map sequential vs parallel tool calling paths for multi-step intent fulfillment.

Constraints

  • MUST express every tool signature in strict JSON Schema syntax with clear docstrings.
  • MUST NOT permit unbounded retries or open-ended parameter exploration by the agent.
  • System instructions MUST enforce input sanitization prior to API dispatch.
  • Keep architecture specifications modular and production-ready for immediate pipeline ingestion.

Output format

A technical specification report organized into:

  1. Executive Architecture Summary (max 200 words)
  2. Tool Schema Registry (full JSON Schema for each tool in {{external_apis}})
  3. Orchestration & Sequence Matrix (sequential vs parallel calling rules)
  4. Security, Privacy & Redaction Directives
  5. Error Fallback and Timeout Decision Trees

Self-review

  • Are all external APIs in {{external_apis}} modeled with complete, valid parameter schemas?
  • Are latency caps from {{latency_budget_ms}} explicitly reflected in timeout handling?
  • Does the redaction protocol fully cover {{sensitive_data_scope}}?
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 engineering10/12 · Adequate

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.

Robustness5/5 · Strong

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-instructions
autonomous-agents-workflows
agent-instructions
tool-calling
orchestration