Skip to main content
Glama

向 SideDesk 人物卡派单

sidedesk_dispatch

Dispatches backend analysis, audit, or coding tasks to a selected persona, running them in an isolated shadow copy that leaves the original project unchanged unless turbo-direct is requested.

Instructions

将后端分析、审计或编码任务派发给 SideDesk 副席中的指定人物卡(自动按人物卡驱动路由至本机 Codex CLI 或 OpenAI 兼容 API)。默认必须在隔离影子副本中执行(read-only / workspace-write),原工程目录零修改;仅当用户明确要求 Turbo / 直接改原目录,或目标人物卡默认权限为 turbo-direct 时,才允许使用 turbo-direct(跳过影子副本与容量熔断,直接在原目录无沙箱执行)。

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "sourceDir": "<sourceDir>",
      "prompt": "<prompt>"
    },
    "name": "sidedesk_dispatch"
  }
}

Input Schema

NameRequiredDescriptionDefault
modelNo可选:覆盖人物卡默认模型
titleNo可选:本话题标题
promptYes清晰、自包含的任务指令与验收标准
sandboxNo权限模式:默认必须优先选 read-only(纯只读分析)或 workspace-write(影子副本隔离读写)。严禁私自升级为 turbo-direct(直接操作原目录无沙箱、无容量熔断);仅当用户明确要求开启 Turbo/直通模式或人物卡默认设为 turbo-direct 时使用。
fastModeNo可选:Codex CLI 闪电加速(2x 速度,会增加订阅额度消耗)
personaIdNo可选:人物卡 ID,默认使用第一张卡(codex-default)
sessionIdNo可选:传入已有副席 sessionId 以续聊同一上下文
sourceDirYes目标工程绝对路径(影子模式下创建隔离副本;turbo-direct 下直接在原目录执行)
waitSecondsNo可选:同步等待秒数(默认 0 立即返回 jobId 供异步查询)
reasoningEffortNo可选:Codex CLI 推理强度(支持 minimal/low/medium/high/xhigh/max/ultra)

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv0.1.0

TDQS

A4.1/5.0
Behavior4/5

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

With no annotations, the description carries the full burden and does substantial work: it explains routing behavior, that the original project directory is left untouched, that turbo-direct skips the shadow copy and capacity circuit breaker and runs unsandboxed, and that fastMode costs extra quota. It stops short of failure modes, permissions/auth needs, and result structure beyond mentioning jobId.

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?

A single front-loaded paragraph that names the action, then the isolation default, then the narrow exception. It is dense but readable, with mild redundancy against the sandbox parameter's own description.

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

Completeness4/5

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

For a ten-parameter, no-output-schema dispatch tool, the description covers the critical decision (sandbox mode) and the safety boundary, and notes that waitSeconds=0 returns a jobId for async lookup. It omits persona selection behavior and error/return details, which the schema only partially covers.

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

Parameters3/5

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

Schema description coverage is 100%, so the schema already documents all ten parameters, including the sandbox prohibition and waitSeconds async semantics. The description largely restates the sandbox policy rather than adding new parameter-level meaning, so the baseline applies.

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?

States a specific verb and resource: dispatching backend analysis, audit, or coding tasks to a named persona card in SideDesk, plus how the card routes work (local Codex CLI or OpenAI-compatible API). This clearly separates it from siblings such as sidedesk_job_status, sidedesk_abort_job, and sidedesk_list_personas.

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

Usage Guidelines4/5

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

Gives an explicit default (must run in an isolated shadow copy with read-only/workspace-write) and the exact conditions for departing from it (user explicitly requests Turbo/direct edit, or the persona's default permission is turbo-direct). It does not name the follow-up sibling for polling, only implying async query via jobId.

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