tianshu-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| TIANSHU_MCP_HOME | No | Path to the .tianshu-mcp directory (e.g. <repository_absolute_path>/.tianshu-mcp). Used to locate acceptance configuration, skills, and other server data. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| prepare_visual_baselineB | 准备视觉基准候选，返回摘要与预览；不采用正式基准。需要用户授权。 |
| approve_visual_baselineA | 仅在用户明确审阅并授权后批准视觉基准。必须核对候选摘要与批准说明；自动返修禁止调用。宿主必须实施实际审批控制。 |
| continue_taskA | 恢复处于 needs_user 的任务。zcode：agent_question 时 message 发往原会话，关闭旧实例/登录/系统权限场景中 message 仅作已处理确认。codex：user_confirmation 时重新接入观察 GUI 内运行（不发送消息）；login_required 时复检环境后重发任务书。 |
| run_taskB | 派活：启动外部 AI-Agent 开发任务并可自动验收返修，异步返回 taskId。ZCode 要求 model=供应商/模型，不支持 mode；TraeWork 的 model 可选并支持 Work/Code/Design mode。task/context 内的 ZCode 项目路径引用会在发送前校验。 |
| query_taskA | 查询任务状态 / 进度 / 最近日志尾部（默认 agent.log 末 40 行）。返回任务 meta 与日志片段。 |
| list_tasksA | 列出历史任务（可按项目路径 / 状态过滤，limit 默认 50）。 |
| get_task_reportB | 取某轮验收报告全文（report.md）。round 缺省取最新一轮。 |
| cancel_taskA | 取消运行中任务：CLI agent 终止进程树；GUI agent（codex 等）尽力点击界面停止按钮并等待 GUI 空闲（有界超时），未确认停止时结果中明示。排队中任务直接移除；终态任务无动作。 |
| verify_taskB | 对已完成任务或项目路径执行一次验收（不改源码）：自动命令检查 + 代码分析（相对 git 基线）。可用 extraChecks 临时加验。需任务/项目二选一。 |
| rework_taskA | 手动返修：把终态任务（failed/needs_attention）重新入队续跑，同一 agent/项目与轮次记账。feedback 为追加指示（建议带上一次验收失败摘要）。 |
| get_profilesA | 查看当前 agent 适配与可执行探测结果（含未安装/调研占位提示）。 |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 11 tools
Most tools have clear boundaries: task lifecycle (run/cancel/continue/rework/query/list/verify) vs. visual baseline (prepare/approve) vs. report/profile. The only mild overlap is verify_task vs. get_task_report (both relate to acceptance), but verify runs a new check while get_task_report reads an existing report.
Tool names mostly follow a verb_noun pattern (get_task_report, prepare_visual_baseline, approve_visual_baseline, cancel_task, continue_task, list_tasks, query_task, rework_task, verify_task). Minor inconsistency: run_task is verb_noun but get_profiles is verb_noun too; all are snake_case and readable. The slight mix of 'task' vs 'tasks' (list_tasks plural vs others singular) is a small deviation.
11 tools is well within the ideal 3-15 range and each tool maps to a distinct operation in the task-management and visual-baseline workflow. The count feels appropriate for the server's purpose: task lifecycle management plus baseline approval.
The task lifecycle is well covered: create (run_task), query (query_task/list_tasks), cancel, continue, rework, verify, and report retrieval. Minor gaps: no explicit delete/archive task tool, and visual baseline has prepare/approve but no explicit reject or list-baselines tool, though these may be intentionally omitted.