Spec Driven Development
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
作者提供的说明(英文原文)
Connect AI assistants to the Tickiti helpdesk API: tickets, queries, reports and admin.
这三项永远不会合并成一个分数:意见、测量和人气是不同的东西。
我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。
选择你使用的 AI 应用,然后复制设置。
作者没有在注册中心列出安装信息。请查看下方作者的 README。
它在你的电脑上运行;安装后,工具列表会显示在你的 AI 应用中。
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月7日
注册名称已通过 GitHub 账户 tickiti 验证,该账户拥有此仓库
我们还没有从 GitHub 读取这个仓库
我们还没有从 GitHub 读取这个仓库
它在你的电脑上运行,没有互联网端点
在你的电脑上运行,改用本地设置
它在你的电脑上运行;我们只能读取互联网端点的工具列表
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
它在 1 项适用的基础检查中通过了 1 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
注册中心没有列出必需的密钥,不过有些服务器会在首次使用时要求你登录。
它在你的电脑上运行。你的 AI 应用会在需要时启动它,数据不会经过作者的服务器(除非它自己调用在线服务)。
GitHub 仓库最后更新于–,最新版本(0.2.0)于33天前进入注册中心。
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。
同一分类中有:Spec Driven Development, Remote Desktop Commander, EmbeddedCI BenchPod
同一分类中其他作者的服务器,按星标数排序。
如何看卡片:★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
Remote Desktop Commander: use files, terminal and processes on your own computers from any AI app.
Drive a BenchPod hardware-in-the-loop bench: power, flash, UART, I2C, CAN, analog and logic.
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
An MCP (Model Context Protocol) server that exposes the Tickiti helpdesk API to AI assistants such as Claude. It is a thin shim over the Tickiti Public API v1 (/api/v1/...): each MCP tool forwards to a v1 endpoint, adding your bearer token and — for writes — an idempotency key. The token's abilities are the security boundary: the server only relays calls, it never widens them, so a read-only token gives a read-only assistant.
📖 Full documentation: https://www.tickiti.com/docs/topic/mcp_server/
The ticket tools have full, validated inputs; the rest of the API is reachable through two general tools, so the whole surface is available without a separate tool per endpoint.
Tickets
| Tool | Ability | Purpose |
|---|---|---|
create_ticket | tickets:write | Open a ticket (subject+content, template, or intervention), with assignee, priority, status and participants; is_public is required |
respond_to_ticket | tickets:write | Post a response; change status, hold, assignee, priority, queue, subject, participants; send_full_email for the whole reply by email |
close_ticket | tickets:write | Close, optionally with a final reply and a resolution |
edit_response | tickets:write | Change a response in place (whole body, find/replace, visibility) with no notification — or publish an internal note and notify |
delete_response | tickets:write | Soft-delete a response |
add_participants / remove_participants | tickets:write | Change participants |
delete_tickets / restore_ticket | tickets:write | Delete tickets (by number, or everything from one sender) and undo it |
merge_tickets | tickets:write | Merge two tickets (the older survives) |
link_tickets / unlink_tickets / list_linked_tickets | tickets:write / read | The Linked panel |
move_responses / copy_responses / split_ticket | tickets:write | Repair a thread: move or copy responses, or split them off into a new linked ticket |
change_originator | tickets:write | Change who raised a ticket |
delete_attachment | tickets:write | Remove one attachment (a malicious file, say), keeping a record of it |
get_ticket / list_responses | tickets:read | A ticket's thread, participants and links; fields / max_body_chars keep long email threads small |
get_response | tickets:read | One response by id |
query_tickets | tickets:read | Search tickets |
query_responses / export_responses | tickets:read | Responses across tickets; counts and group-by tallies; NDJSON export to disk |
get_attachment / download_attachment | tickets:read | Attachment bytes (download streams any size to a file) |
list_queues / list_teams / list_staff / list_resolutions / get_search_modes | tickets:read | Queue names and settings, team keys, staff addresses, resolution categories, the search vocabulary |
Settings, templates, mail, administration
| Tool | Ability | Purpose |
|---|---|---|
list_perspectives / get_perspective | settings:read | Saved views, with conditions and sort orders |
list_watchlists | settings:read | Watchlists |
list_stock_responses / get_stock_response | settings:read | Stock responses (search, bodies) |
create_stock_response / update_stock_response / delete_stock_response | settings:write | Edit stock responses |
get_template / search_templates / render_template | templates:read | Templates and FAQs; render one against data without sending |
create_template / update_template | templates:write | Edit templates (updates change only the fields given) |
upload_template_image | templates:write / settings:write | Upload an image for a template or stock response |
send_email / search_sent_mail | mail:write / mail:read | Send mail; search what was sent |
list_api_tokens / create_api_token / update_api_token / revoke_api_token | administration:* | API keys (a token only grants abilities it holds) |
list_workflow | workflow:read | Resolution categories, interventions, escalations |
run_report | reports:read | Analytics reports |
get_token_info | any | What this token can do: abilities, owner roles, plan |
list_instances | — | The Tickiti instances configured (see below) |
list_endpoints | — | Every API endpoint, with abilities, path and body parameters |
tickiti_call | per endpoint | Call any /api/v1 endpoint by family and action |
For anything beyond the named tools (mail, templates, workflow writes, administration, supervisor), the assistant uses list_endpoints to discover the action, then tickiti_call to run it — covering all of the v1 API.
git clone https://github.com/tickiti/tickiti-mcp.git
cd tickiti-mcp npm install npm run build
The built server is dist/server.js.
The server reads two environment variables (it fails fast on startup if either is missing):
| Variable | Purpose |
|---|---|
TICKITI_API_BASE | Your Tickiti install's public address, no trailing slash — e.g. https://support.example.com. The server appends /api/v1/…. |
TICKITI_API_TOKEN | The bearer token. Its abilities determine what the assistant can do. |
To reach several Tickiti installs from one server (production and staging, say), add TICKITI_INSTANCES: JSON, or the path of a JSON file, mapping a name to { "base", "token" }. The TICKITI_API_BASE / TICKITI_API_TOKEN pair stays the default instance, named by TICKITI_INSTANCE_NAME (default default).
TICKITI_INSTANCE_NAME=production
TICKITI_INSTANCES='{"staging": {"base": "https://staging.example.com", "token": "…"}}'
Every tool then takes an optional instance argument; list_instances shows the names.
claude mcp add tickiti \
--env TICKITI_API_BASE=https://support.example.com \ --env TICKITI_API_TOKEN=YOUR_TICKITI_API_TOKEN \ -- node /absolute/path/to/tickiti-mcp/dist/server.js
Confirm with claude mcp list (or /mcp in a session). Remove with claude mcp remove tickiti.
Other MCP clients configure servers in their own settings file, but the shape is the same: run node /absolute/path/to/tickiti-mcp/dist/server.js as a stdio server with TICKITI_API_BASE and TICKITI_API_TOKEN set in its environment.
The server adds no permissions of its own. Every call runs as the staff user the token belongs to, gated by the token's abilities — exactly as a direct API call would be. To limit what an assistant can do, mint a narrowly-scoped token:
tickets:read, reports:read) gives an assistant that can look but not change anything.The ticket-writing tools send an idempotency key with every call, so a retried request never creates a duplicate ticket or response.
Long bodies can be given as content_path (a local file) instead of content, and files and images always go by path: the server reads them, so their bytes never pass through the model.
| File | Role |
|---|---|
src/client.ts | Request core: base URL, bearer auth, idempotency, error normalisation |
src/result.ts | Maps an API result into the MCP tool-result envelope |
src/manifest.ts | Helpers over the generated route manifest (lookup, path building) |
src/generated/manifest.ts | Auto-generated route table (do not edit) |
src/tools/tickets.ts | Ticket writes — verified input schemas |
src/tools/ticket-reads.ts | Ticket, response and attachment reads |
src/tools/ticket-operations.ts | Delete, merge, link, move, originator, attachment delete, lookups |
src/tools/settings-writes.ts | Stock responses and templates |
src/tools/admin.ts | Mail, template rendering and images, perspectives, API keys, instances |
src/tools/reads.ts | Named read tools (settings / workflow / reports) |
src/tools/generic.ts | list_endpoints + tickiti_call |
src/param-hints.ts | Body parameters list_endpoints shows for endpoints without a dedicated tool |
src/server.ts | Entry point: registers tools, connects the stdio transport |
scripts/build-manifest.mjs | Regenerates the manifest from the Tickiti route table |
src/generated/manifest.ts is generated from Tickiti's own route table (php artisan route:list --json), so abilities, roles, plan gates and path params are never hand-maintained. Regenerate against a Tickiti checkout after the API changes:
TICKITI_DIR=/path/to/tickiti npm run manifest
There is an end-to-end sweep over every endpoint in tests/all-paths.mjs (npm run test:paths, needs a base URL and a full-ability token against a scratch instance), and an offline check of the tool registration in tests/smoke-tools.mjs (npm run build && node tests/smoke-tools.mjs).
Several tools call endpoints added to Tickiti in October 2026. Against an older install they return 404; list_endpoints describes the newest API.
MIT © Oxenic