job-mcp
job-mcp
FastMCP-сервер (транспорт stdio) с:
Ресурс
cv://profile— ваше резюме (CV) в формате JSON, которое может прочитать LLM-клиент.Инструмент
send_email— отправляет письмо по SMTP с помощью Red Mail.
Рабочий процесс: вы передаёте MCP-клиенту в чате вакансию + email рекрутера → клиент
читает cv://profile → составляет адаптированное сопроводительное письмо → вызывает
send_email для его отправки.
1. Установка
cd job-mcp
python -m venv .venv
source .venv/bin/activate # Windows: .venv\Scripts\activate
pip install -e .Related MCP server: MCP Email Server
2. Настройка учётных данных SMTP
cp .env.example .envОтредактируйте .env, указав ваш SMTP-хост и учётные данные. Для Gmail используйте
пароль приложения (не обычный пароль от аккаунта)
и сначала включите двухэтапную проверку.
3. Заполнение своего резюме
Отредактируйте src/job_mcp/data/cv.json, указав свои реальные данные (контакты, навыки, опыт,
образование, проекты, сертификаты). Ресурс просто читает этот файл как есть.
4. Запуск сервера отдельно (проверка работоспособности)
python -m job_mcp.serverОн будет находиться в ожидании MCP JSON-RPC-сообщений через stdin/stdout (это нормально — он предназначен для запуска MCP-клиентом, а не для интерактивного использования).
Вы также можете просмотреть его с помощью MCP dev inspector:
mcp dev src/job_mcp/server.py5. Подключение к MCP-клиенту
Пример конфигурации для клиента, который запускает серверы через stdio (например,
claude_desktop_config.json для Claude Desktop):
{
"mcpServers": {
"job-mcp": {
"command": "/absolute/path/to/job-mcp/.venv/bin/python",
"args": ["-m", "job_mcp.server"],
"env": {
"SMTP_HOST": "smtp.gmail.com",
"SMTP_PORT": "587",
"SMTP_USERNAME": "you@gmail.com",
"SMTP_PASSWORD": "your-app-password",
"SMTP_SENDER_EMAIL": "you@gmail.com",
"SMTP_SENDER_NAME": "Your Name"
}
}
}
}(Вы можете указать учётные данные в env здесь, а не в .env — работают оба варианта,
поскольку config.py в любом случае читает их из os.environ.)
6. Использование
В клиенте вставьте что-то вроде:
Вот вакансия и email рекрутера. Прочитай мой CV-ресурс, затем составь и отправь адаптированное сопроводительное письмо.
Вакансия: <вставьте текст> Email рекрутера: hr@company.com
Клиент должен вызвать cv://profile, составить письмо, используя оба фрагмента контекста,
и затем вызвать send_email(to=..., subject=..., body_html=..., body_text=...).
Примечания / что стоит укрепить позже
Сейчас в
send_emailнет встроенного в сам сервер шага подтверждения — если вы хотите, чтобы защитный механизм «проверка перед отправкой» обеспечивался на стороне сервера (а не только на стороне клиента), добавьте двухэтапный процесс (например, инструментdraft_email, который возвращает предпросмотр + id, и инструментconfirm_send, который непосредственно отправляет письмо).Рассмотрите ограничение частоты (rate limiting) или логирование отправленных писем, чтобы избежать случайных повторных отправок.
cv.jsonчитается заново при каждом обращении к ресурсу, поэтому вы можете редактировать его без перезапуска сервера.
Available Tools
1 toolsend_emailA
Send an email via SMTP using Red Mail. Use this to send the final, tailored job-application email to the recruiter/HR address. Provide a polished HTML body (and optionally a plain-text fallback).
| Name | Required | Description | Default |
|---|---|---|---|
| cc | No | ||
| to | Yes | ||
| bcc | No | ||
| subject | Yes | ||
| reply_to | No | ||
| body_html | Yes | ||
| body_text | No |
TDQS
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Since no annotations are provided, the description carries the behavioral disclosure burden. It transparently states that the tool sends an email via SMTP, which is an external side effect, and adds a practical requirement to provide polished HTML. It omits details like SMTP prerequisites or failure modes, but the core behavior is explicit.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is two sentences long, front-loaded with the core action, and every clause adds value: the action, the use case, and the body-format expectation. No redundant or generic phrases.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
The tool has 7 parameters, no annotations, and no output schema, so the description must be reasonably self-contained. It covers the main action, the intended purpose, and the key body requirement. However, it leaves several parameter semantics and any error/response behavior unaddressed. The operation is simple and the names are fairly self-evident, so the gaps are moderate rather than severe.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 0%, so the description must compensate. It only explains body_html and optionally body_text; to, subject, cc, bcc, and reply_to are left undocumented. While the parameter names are conventional, the description does not add sufficient meaning for a zero-coverage schema, especially for required fields like 'to' and 'subject'.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the action ('Send an email via SMTP using Red Mail') and the specific use case ('final, tailored job-application email to the recruiter/HR address'). It is a precise verb+resource combination, with no ambiguity or tautology.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description explicitly says 'Use this to send the final, tailored job-application email to the recruiter/HR address', giving a clear when-to-use context. There are no sibling tools listed, so there is no opportunity to name alternatives or when-not scenarios, but the intended scenario is well defined.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
1 tool update
v0.1.0- First observed
send_email
TDQS
Scored across 1 tool
With only one tool, there is no possibility of ambiguity or misselection. The tool's purpose is clearly distinct simply because it is the only tool.
A single tool name cannot be inconsistent with itself. 'send_email' follows a clear verb_noun convention, and there are no other names to compare against.
One tool feels too thin for a server named 'job-mcp', which implies a broader job-application workflow. While sending an email is a concrete task, the server likely needs supporting tools for content generation or attachment handling to be useful.
The surface is severely limited: it only sends an email. Missing operations include email preview, attachment support, and any integration with job-application data, leaving obvious gaps for real-world workflows.
Maintenance
Related MCP Connectors
Auto-apply to jobs: matches your CV, tailors a fresh CV per posting, and applies for you.
Send transactional emails for AI agents via SMTP. Templates included.
Litjob AI job application assistant: job matches, tailored CVs and cover letters. You submit.
Tailor resumes, score ATS fit, find hiring leads, and send cold outreach, from your agent.
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceAutomates cold email outreach for job applications by parsing job postings, generating personalized emails using AI, and sending them or saving as drafts in Gmail with resume attachments.1-
- FlicenseNot gradedqualityDmaintenanceEnables LLMs to read, send, and manage emails via POP3 and SMTP with TLS support.6-
- AlicenseNot gradedqualityDmaintenanceEnables LLMs to access candidate information including resume, LinkedIn, GitHub, and contact via email.14 npm81MIT
- AlicenseAqualityAmaintenanceFree MCP and skills for remote and overseas job search across Himalayas, Jobicy, Arbeitnow and company ATS boards, with source links and eligibility checks. Optional local drafts, authorized email sending and reply tracking. Hosted MCP: https://mcp.worldrole.work/mcp — free anonymous access to 20 distinct jobs per network per UTC day, or 50 per free account; other request limits apply.914MIT