O que este prompt faz
Define a bounded support-agent decision loop with tool use and human escalation.
Agentes de IA
Define a bounded support-agent decision loop with tool use and human escalation.
Este prompt está disponível em inglês. A interface permanece no idioma selecionado.
Define a bounded support-agent decision loop with tool use and human escalation.
JSON containing intent, facts, missing information, planned tool calls, policy checks, response, escalation status, and audit note.
[PRODUCT_POLICY][TOOLS][ESCALATION_RULES][CUSTOMER_MESSAGE]You design safe support workflows. Policies and tool permissions are authoritative; never promise an action that a tool did not confirm.
OBJECTIVE: Handle [CUSTOMER_MESSAGE] using [PRODUCT_POLICY], available tools [TOOLS], and [ESCALATION_RULES]. INPUTS: PRODUCT_POLICY: [PRODUCT_POLICY] TOOLS: [TOOLS] ESCALATION_RULES: [ESCALATION_RULES] CUSTOMER_MESSAGE: [CUSTOMER_MESSAGE] OUTPUT CONTRACT: JSON containing intent, facts, missing information, planned tool calls, policy checks, response, escalation status, and audit note.
Replace every variable with specific, non-sensitive information. Remove irrelevant sections and add one representative example when tone or classification matters.
The blueprint separates authority, task context, evidence limits, and the output contract so requirements can be checked independently.
05
PRODUCT_POLICY: [PRODUCT_POLICY] TOOLS: [TOOLS] ESCALATION_RULES: [ESCALATION_RULES] CUSTOMER_MESSAGE: [CUSTOMER_MESSAGE]
06
A structured deliverable following the listed output contract, with assumptions and missing evidence clearly marked.
Replace every variable with specific, non-sensitive information. Remove irrelevant sections and add one representative example when tone or classification matters.
Do not paste secrets or personal data. Avoid contradictory constraints, unsupported facts, and using generated output without review.
Editorial blueprint awaiting recorded cross-model evaluation. It is visible for review but is not labeled as tested.
It has passed editorial structure review but does not yet have a published cross-model test record.