WACrew

caixa_de_entrada

Shows which clients are waiting for a reply and from which project across collected WhatsApp chats, listing pending requests first and oldest waits on top.

Instructions

Quem está esperando resposta e de qual projeto, em todos os chats coletados (o operador não precisa abrir o WhatsApp). Use pra 'quem me mandou mensagem?', 'tem cliente esperando?', 'o que rolou hoje?'. Cada item traz o chat, o projeto (cliente, pasta do repo e GitHub, quando o chat está linkado) e, se alguém espera, esperando (quem, prévia, desde quando, há quanto tempo); pendentes vêm primeiro, quem espera há mais tempo no topo. Esperando = a última fala é de cliente (ou te mencionaram) e ninguém da conta respondeu depois; chat silenciado ou marcado resolvido não entra. Só lê: não marca nada como visto (ao contrário de novidades). todos: true inclui os chats em que cliente escreveu desde desde mesmo já respondidos, com do_cliente_desde. Conversa que o cliente fechou ("ok", "combinado", "valeu", figurinha) não entra: vem só a contagem em encerrados. Espera cuja última fala do cliente passou de dias (default 7) vem resumida em antigos (nome e tempo). Ao responder o operador, agrupe por projeto, diga em uma linha o que cada um quer e ainda descarte o que for só social; cite os antigos numa linha. Num repositório ligado a um chat (.claude/whatsapp.json, bloco wa-link no CLAUDE.md), passe grupos com o(s) slug(s) dele em vez de ler a caixa inteira; pra 'teve mensagem nova?' esta tool basta, e ler_mensagens (bem mais caro) só entra se precisar do conteúdo. Se vier avisos (serviço de IA falhando), mostre ANTES de tudo, numa linha cada. Mostre a data e hora de Brasília da última mensagem (ultima_brt) com quanto tempo faz em relação a agora_brt (ex: "28/10 09:34, há 2h"). Pra ler a conversa, ler_mensagens(grupo).

Tool call template

This tool declares no required parameters. Add optional parameters from the input schema as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {},
    "name": "caixa_de_entrada"
  }
}

Input Schema

Name	Required	Description
`dias`	No	default 7: espera mais velha que isso vai pra `antigos`
`desde`	No	ISO (ex 2026-10-28T09:00:00-03:00); só no modo todos; default = últimas 24 h
`todos`	No	default false = só quem espera resposta; true = também o que se mexeu desde `desde`
`grupos`	No	slugs pra restringir

Schema Changelog

Changes observed during successful MCP inspections.

First observedv0.1.0Nov 14, 2026

TDQS

A4.8/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

There are no annotations, so the description has to carry the full burden, and it does: it is strictly read-only, and unlike `novidades` it does not mark anything as seen. It also defines the exact semantics of `esperando` (the last message is from the client or mentions the operator, and nobody on the account replied afterwards), excludes muted or resolved chats, excludes chats that the client closed (with greetings or stickers), and reports those exclusions only as a count in `encerrados`.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

It is dense but almost every sentence carries real information, and the operational definition of `esperando` is front-loaded. The only slight bloat is the embedded output-formatting coaching (grouping by project, citing `antigos` on one line, the '28/10 09:34, há 2h' example), which drifts from defining the tool into prescribing response style.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

With no output schema, the description must explain the return shape, and it does: each item carries the chat, `projeto` (client, repo folder, and GitHub when linked), and `esperando` (who, preview, since, and how long), plus `encerrados`, `antigos`, `avisos`, and `ultima_brt`. Ordering rules (pending first, longest waiting at the top) and precedence for `avisos` are also specified, so nothing an agent needs in order to call the tool correctly is missing.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is already 100%, so the baseline is 3, but the description still adds meaning beyond the schema: `todos: true` is tied to the `do_cliente_desde` field, `dias` is tied to the `antigos` rollup, and `grupos` is tied to slugs from a repo link (.claude/whatsapp.json / wa-link). It tells the agent why these parameters exist, not just what they are.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description opens with a specific verb-and-resource statement: it reports who is waiting for a reply and from which project, across all collected chats, without opening WhatsApp. It explicitly distinguishes itself from `novidades` (this one doesn't mark anything as seen) and from `ler_mensagens` (which is far more expensive and only needed when the content itself is required).

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It gives concrete trigger phrases ('quem me mandou mensagem?', 'tem cliente esperando?', 'o que rolou hoje?'), names the alternative tool and the condition that selects it ('ler_mensagens só entra se precisar do conteúdo'), and explains when to scope with `grupos` instead of reading the whole inbox. Nothing is left to inference.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.