visa-jobs-mcp 是一个MCP服务器，可帮助代理从新的LinkedIn搜索中找到相关工作，并具有可选的签证感知过滤功能。

它是为速度和实际结果而构建的：

在某个位置搜索工作，
将雇主与赞助历史进行匹配，
返回带有链接和联系信息的可操作结果，
将用户数据保存在本地。

开始使用

1.在macOS上安装（Homebrew）

brew tap neosh11/visa-jobs-mcp
brew install neosh11/visa-jobs-mcp/visa-jobs-mcp

2.在macOS/Linux上安装（bash安装程序）

curl -fsSL https://raw.githubusercontent.com/neosh11/visa-jobs-mcp/main/scripts/install.sh | bash

固定特定版本：

curl -fsSL https://raw.githubusercontent.com/neosh11/visa-jobs-mcp/main/scripts/install.sh | bash -s -- --version 0.3.1

3.在Windows上安装（PowerShell安装程序）

irm https://raw.githubusercontent.com/neosh11/visa-jobs-mcp/main/scripts/install.ps1 | iex

固定特定版本：

$env:VISA_JOBS_MCP_VERSION = "0.3.1"
irm https://raw.githubusercontent.com/neosh11/visa-jobs-mcp/main/scripts/install.ps1 | iex

4.在Codex（macOS/Linux）中注册

codex mcp add visa-jobs-mcp --env VISA_JOB_SITES=linkedin -- visa-jobs-mcp

在Codex（Windows PowerShell）中注册：

codex mcp add visa-jobs-mcp --env VISA_JOB_SITES=linkedin -- "$env:LOCALAPPDATA\\Programs\\visa-jobs-mcp\\bin\\visa-jobs-mcp.exe"

验证：

codex mcp list
codex mcp get visa-jobs-mcp

5.在聊天中使用它

在新的食典会议上，自然地问：

Find IT jobs in Delhi for my skills.
Set my visa preference to E3 and find software engineer jobs in New York that sponsor E3.

从源代码构建（可选）

go build -o visa-jobs-mcp ./cmd/visa-jobs-mcp
./visa-jobs-mcp --version

如果你需要刷新 data/companies.csv 从源代码运行：

./scripts/run_internal_pipeline.sh

注意：MCP运行时仅支持Go。Python只需要用于运行内部数据集管道的维护人员。

实时LinkedIn E2E（手册）

对LinkedIn进行真正的端到端搜索（不是简单的测试）：

./scripts/run_live_linkedin_e2e.sh

可选超时覆盖：

VISA_E2E_TEST_TIMEOUT=8m ./scripts/run_live_linkedin_e2e.sh

可选现场测试参数：

VISA_E2E_VISA_TYPE=E3 \
VISA_E2E_LOCATION="New York, NY" \
VISA_E2E_JOB_TITLE="Software Engineer" \
./scripts/run_live_linkedin_e2e.sh

它支持什么

仅限LinkedIn搜索。
一般技能/地点求职，无需签证设置。
可选的签证偏好匹配 preferred_visa_types 已设置。
支持分页和恢复的搜索会话。
保存的作业和忽略的作业。
雇主联系人提取（如有）。
本地第一私有数据存储。
不使用代理。
MCP运行时内没有LLM调用（代理处理推理）。

核心MCP工具

start_job_search
get_job_search_status
get_job_search_results
set_user_preferences
start_visa_job_search
get_visa_job_search_status
get_visa_job_search_results
save_job_for_later
ignore_job
list_saved_jobs
list_ignored_jobs
get_mcp_capabilities

提示：让客服打电话 get_mcp_capabilities 首先是机器可读的合同。

MCP合同（已生成）

Expand MCP Contract (auto-generated)

生成自 get_mcp_capabilities() 通过 scripts/generate_contract_docs.py.

服务器

server: visa-jobs-mcp
version: 0.3.1
capabilities_schema_version: 1.2.0
confidence_model_version: v1.1.0-rules-go

搜索前需要

tool: start_job_search
required_fields: user_id

设计决策

agent_is_reasoning_layer: True
background_search_runs_local_persistence: True
data_not_shared_or_sold: True
first_class_job_management: True
free_forever: True
fresh_job_search_per_query: True
ignored_companies_local_persistence: True
ignored_jobs_local_persistence: True
license: MIT
llm_api_keys_required_by_mcp: False
llm_runtime_inside_mcp: False
no_fake_reviews_or_bot_marketing: True
proxies_used: False
rate_limit_backoff_retries: True
saved_jobs_local_persistence: True
search_sessions_local_persistence: True
strict_user_visa_match: False
strictness_modes_supported: ['balanced', 'strict']
supported_job_sites: ['linkedin']
visa_matching_optional: True

默认值

dataset_stale_after_days: 30
job_db_path: data/app/visa_jobs.db
max_scan_results: 1200
max_search_sessions_per_user: 20
rate_limit_initial_backoff_seconds: 2
rate_limit_max_backoff_seconds: 30
rate_limit_retry_window_seconds: 180
scan_multiplier: 8
search_run_ttl_seconds: 21600
search_session_ttl_seconds: 21600
strictness_mode: strict
tool_call_soft_timeout_seconds: 48

工具

工具	描述	必需输入	可选输入
`get_mcp_capabilities`	返回MCP功能、工具和代理自我发现合同。	-	-
`set_user_preferences`	保存用户的签证偏好，以便进行可选的签证特定匹配。	`user_id`, `preferred_visa_types`	-
`set_user_constraints`	保存用于个性化指导的紧迫性和工作模式限制。	`user_id`	-
`get_user_preferences`	获取已保存的用户首选项和约束。	`user_id`	-
`get_user_readiness`	报告用户和本地数据集是否已准备好进行搜索。	`user_id`	-
`find_related_titles`	返回相邻的角色头衔以扩大低收益搜索。	`job_title`	-
`add_user_memory_line`	添加一条个人资料记忆行（技能、目标、恐惧、限制）。	`user_id`, `content`	-
`query_user_memory_blob`	使用可选的文本过滤查询用户的本地内存blob	`user_id`	-
`delete_user_memory_line`	从本地blob中按id删除一行内存。	`user_id`, `line_id`	-
`save_job_for_later`	将作业保存到用户的本地候选名单中以供跟进。	`user_id`	`job_url`, `result_id`, `session_id`
`list_saved_jobs`	按逆时间顺序列出已保存的作业。	`user_id`	-
`delete_saved_job`	从本地候选名单中删除一个已保存的作业。	`user_id`, `saved_job_id`	-
`ignore_job`	从此用户的未来结果中隐藏一个作业。	`user_id`	`job_url`, `result_id`, `session_id`
`list_ignored_jobs`	按逆时间顺序列出忽略的作业。	`user_id`	-
`unignore_job`	按id取消隐藏以前忽略的作业	`user_id`, `ignored_job_id`	-
`ignore_company`	在以后的搜索中隐藏公司的所有工作。	`user_id`	-
`list_ignored_companies`	按逆时间顺序列出被忽略的公司。	`user_id`	-
`unignore_company`	从忽略列表中删除一家公司。	`user_id`, `ignored_company_id`	-
`mark_job_applied`	将作业标记为已应用并保持管道状态。	`user_id`	-
`update_job_stage`	更新跟踪作业的生命周期阶段（保存/申请/面试等）。	`user_id`, `stage`	-
`list_jobs_by_stage`	列出按生命周期阶段筛选的跟踪作业。	`user_id`, `stage`	-
`add_job_note`	在跟踪的作业记录中附加或追加注释。	`user_id`, `note`	-
`list_recent_job_events`	列出最近的阶段转换和生命周期事件。	`user_id`	-
`get_job_pipeline_summary`	按阶段汇总一个用户的跟踪管道计数。	`user_id`	-
`clear_search_session`	删除用户的一个缓存搜索会话或所有会话。	`user_id`	-
`export_user_data`	跨商店导出用户的所有本地记录。	`user_id`	-
`delete_user_data`	永久删除用户的所有本地记录。	`user_id`, `confirm`	-
`get_best_contact_strategy`	为工作推荐最佳的外联渠道/联系人。	`user_id`	-
`generate_outreach_message`	根据用户和角色制定一份实用的外联草案。	`user_id`	-
`start_job_search`	无需签证偏好即可开始后台求职。	`location`, `job_title`, `user_id`	-
`get_job_search_status`	轮询后台作业搜索运行的增量进度/事件。	`user_id`, `run_id`	-
`get_job_search_results`	从后台作业搜索运行中获取当前结果页面。	`user_id`, `run_id`	-
`cancel_job_search`	请求取消正在进行的后台作业搜索运行。	`user_id`, `run_id`	-
`start_visa_job_search`	启动长扫描的后台搜索运行。	`location`, `job_title`, `user_id`	-
`get_visa_job_search_status`	轮询后台搜索运行的增量进度/事件。	`user_id`, `run_id`	-
`get_visa_job_search_results`	从后台搜索运行中获取当前结果页面。	`user_id`, `run_id`	-
`cancel_visa_job_search`	请求取消正在进行的后台运行。	`user_id`, `run_id`	-
`discover_latest_dol_disclosure_urls`	发现最新的DOL LCA/PERM披露来源。	-	-
`run_internal_dol_pipeline`	运行内部管道以刷新赞助商公司数据集。	-	-
`refresh_company_dataset_cache`	清除并重新加载内存中的公司数据集缓存。	-	-

搜索响应字段

run
status
stats
guidance
dataset_freshness
pagination
recovery_suggestions
jobs[].result_id
jobs[].job_url
jobs[].title
jobs[].company
jobs[].location
jobs[].site
jobs[].date_posted
jobs[].description_fetched
jobs[].description
jobs[].description_excerpt
jobs[].salary_text
jobs[].salary_currency
jobs[].salary_interval
jobs[].salary_min_amount
jobs[].salary_max_amount
jobs[].salary_source
jobs[].job_type
jobs[].job_level
jobs[].company_industry
jobs[].job_function
jobs[].job_url_direct
jobs[].is_remote
jobs[].employer_contacts
jobs[].visa_counts
jobs[].visas_sponsored
jobs[].visa_match_strength
jobs[].eligibility_reasons
jobs[].confidence_score
jobs[].confidence_model_version
jobs[].agent_guidance

路径

dataset_default: data/companies.csv
ignored_companies_default: data/config/ignored_companies.json
ignored_jobs_default: data/config/ignored_jobs.json
job_management_db_default: data/app/visa_jobs.db
pipeline_manifest_default: data/pipeline/last_run.json
saved_jobs_default: data/config/saved_jobs.json
search_runs_store_default: data/config/search_runs.json
search_session_store_default: data/config/search_sessions.json
user_memory_blob_default: data/config/user_memory_blob.json
user_preferences_default: data/config/user_preferences.json

反对

build_company_dataset_from_dol_disclosures -> run_internal_dol_pipeline (soft_deprecated)

Raw Capabilities JSON

{
  "capabilities_schema_version": "1.2.0",
  "confidence_model_version": "v1.1.0-rules-go",
  "defaults": {
    "dataset_stale_after_days": 30,
    "job_db_path": "data/app/visa_jobs.db",
    "max_scan_results": 1200,
    "max_search_sessions_per_user": 20,
    "rate_limit_initial_backoff_seconds": 2,
    "rate_limit_max_backoff_seconds": 30,
    "rate_limit_retry_window_seconds": 180,
    "scan_multiplier": 8,
    "search_run_ttl_seconds": 21600,
    "search_session_ttl_seconds": 21600,
    "strictness_mode": "strict",
    "tool_call_soft_timeout_seconds": 48
  },
  "deprecations": [
    {
      "name": "build_company_dataset_from_dol_disclosures",
      "replacement": "run_internal_dol_pipeline",
      "status": "soft_deprecated"
    }
  ],
  "design_decisions": {
    "agent_is_reasoning_layer": true,
    "background_search_runs_local_persistence": true,
    "data_not_shared_or_sold": true,
    "first_class_job_management": true,
    "free_forever": true,
    "fresh_job_search_per_query": true,
    "ignored_companies_local_persistence": true,
    "ignored_jobs_local_persistence": true,
    "license": "MIT",
    "llm_api_keys_required_by_mcp": false,
    "llm_runtime_inside_mcp": false,
    "no_fake_reviews_or_bot_marketing": true,
    "proxies_used": false,
    "rate_limit_backoff_retries": true,
    "saved_jobs_local_persistence": true,
    "search_sessions_local_persistence": true,
    "strict_user_visa_match": false,
    "strictness_modes_supported": [
      "balanced",
      "strict"
    ],
    "supported_job_sites": [
      "linkedin"
    ],
    "visa_matching_optional": true
  },
  "pagination_contract": {
    "next_step": "use pagination.next_offset to request the next page",
    "offset_model": "offset is applied to accepted jobs, not raw scraped jobs",
    "result_id_aliases": "use jobs[].result_id in save_job_for_later/ignore_job to avoid URL copy friction",
    "scan_behavior": "server can increase raw scan depth when auto_expand_scan=true",
    "session_behavior": "pass search_session.session_id for stable paging without redundant rescans"
  },
  "paths": {
    "dataset_default": "data/companies.csv",
    "ignored_companies_default": "data/config/ignored_companies.json",
    "ignored_jobs_default": "data/config/ignored_jobs.json",
    "job_management_db_default": "data/app/visa_jobs.db",
    "pipeline_manifest_default": "data/pipeline/last_run.json",
    "saved_jobs_default": "data/config/saved_jobs.json",
    "search_runs_store_default": "data/config/search_runs.json",
    "search_session_store_default": "data/config/search_sessions.json",
    "user_memory_blob_default": "data/config/user_memory_blob.json",
    "user_preferences_default": "data/config/user_preferences.json"
  },
  "rate_limit_contract": {
    "failure_message": "asks agent to retry shortly when the retry window is exhausted",
    "max_retry_window_seconds": 180,
    "retry_behavior": "automatic exponential backoff on rate-limit errors (429/Too Many Requests)"
  },
  "required_before_search": {
    "required_fields": [
      "user_id"
    ],
    "tool": "start_job_search"
  },
  "search_response_fields_for_agents": [
    "run",
    "status",
    "stats",
    "guidance",
    "dataset_freshness",
    "pagination",
    "recovery_suggestions",
    "jobs[].result_id",
    "jobs[].job_url",
    "jobs[].title",
    "jobs[].company",
    "jobs[].location",
    "jobs[].site",
    "jobs[].date_posted",
    "jobs[].description_fetched",
    "jobs[].description",
    "jobs[].description_excerpt",
    "jobs[].salary_text",
    "jobs[].salary_currency",
    "jobs[].salary_interval",
    "jobs[].salary_min_amount",
    "jobs[].salary_max_amount",
    "jobs[].salary_source",
    "jobs[].job_type",
    "jobs[].job_level",
    "jobs[].company_industry",
    "jobs[].job_function",
    "jobs[].job_url_direct",
    "jobs[].is_remote",
    "jobs[].employer_contacts",
    "jobs[].visa_counts",
    "jobs[].visas_sponsored",
    "jobs[].visa_match_strength",
    "jobs[].eligibility_reasons",
    "jobs[].confidence_score",
    "jobs[].confidence_model_version",
    "jobs[].agent_guidance"
  ],
  "server": "visa-jobs-mcp",
  "tools": [
    {
      "description": "Return MCP capabilities, tools, and contracts for agent self-discovery.",
      "name": "get_mcp_capabilities",
      "required_inputs": []
    },
    {
      "description": "Save the user's visa preferences for optional visa-specific matching.",
      "name": "set_user_preferences",
      "required_inputs": [
        "user_id",
        "preferred_visa_types"
      ]
    },
    {
      "description": "Save urgency and work-mode constraints used for personalized guidance.",
      "name": "set_user_constraints",
      "required_inputs": [
        "user_id"
      ]
    },
    {
      "description": "Fetch the saved user preferences and constraints.",
      "name": "get_user_preferences",
      "required_inputs": [
        "user_id"
      ]
    },
    {
      "description": "Report whether the user and local dataset are ready for search.",
      "name": "get_user_readiness",
      "required_inputs": [
        "user_id"
      ]
    },
    {
      "description": "Return adjacent role titles to widen low-yield searches.",
      "name": "find_related_titles",
      "required_inputs": [
        "job_title"
      ]
    },
    {
      "description": "Append a profile memory line (skills, goals, fears, constraints).",
      "name": "add_user_memory_line",
      "required_inputs": [
        "user_id",
        "content"
      ]
    },
    {
      "description": "Query the user's local memory blob with optional text filtering.",
      "name": "query_user_memory_blob",
      "required_inputs": [
        "user_id"
      ]
    },
    {
      "description": "Delete one memory line by id from the local blob.",
      "name": "delete_user_memory_line",
      "required_inputs": [
        "user_id",
        "line_id"
      ]
    },
    {
      "description": "Save a job to the user's local shortlist for follow-up.",
      "name": "save_job_for_later",
      "optional_inputs": [
        "job_url",
        "result_id",
        "session_id"
      ],
      "required_inputs": [
        "user_id"
      ]
    },
    {
      "description": "List saved jobs in reverse-chronological order.",
      "name": "list_saved_jobs",
      "required_inputs": [
        "user_id"
      ]
    },
    {
      "description": "Remove one saved job from the local shortlist.",
      "name": "delete_saved_job",
      "required_inputs": [
        "user_id",
        "saved_job_id"
      ]
    },
    {
      "description": "Hide one job from future results for this user.",
      "name": "ignore_job",
      "optional_inputs": [
        "job_url",
        "result_id",
        "session_id"
      ],
      "required_inputs": [
        "user_id"
      ]
    },
    {
      "description": "List ignored jobs in reverse-chronological order.",
      "name": "list_ignored_jobs",
      "required_inputs": [
        "user_id"
      ]
    },
    {
      "description": "Unhide a previously ignored job by id.",
      "name": "unignore_job",
      "required_inputs": [
        "user_id",
        "ignored_job_id"
      ]
    },
    {
      "description": "Hide all jobs from a company in future searches.",
      "name": "ignore_company",
      "required_inputs": [
        "user_id"
      ]
    },
    {
      "description": "List ignored companies in reverse-chronological order.",
      "name": "list_ignored_companies",
      "required_inputs": [
        "user_id"
      ]
    },
    {
      "description": "Remove one company from the ignored list.",
      "name": "unignore_company",
      "required_inputs": [
        "user_id",
        "ignored_company_id"
      ]
    },
    {
      "description": "Mark a job as applied and persist pipeline state.",
      "name": "mark_job_applied",
      "required_inputs": [
        "user_id"
      ]
    },
    {
      "description": "Update lifecycle stage for a tracked job (saved/applied/interview/etc).",
      "name": "update_job_stage",
      "required_inputs": [
        "user_id",
        "stage"
      ]
    },
    {
      "description": "List tracked jobs filtered by lifecycle stage.",
      "name": "list_jobs_by_stage",
      "required_inputs": [
        "user_id",
        "stage"
      ]
    },
    {
      "description": "Attach or append a note to a tracked job record.",
      "name": "add_job_note",
      "required_inputs": [
        "user_id",
        "note"
      ]
    },
    {
      "description": "List recent stage transitions and lifecycle events.",
      "name": "list_recent_job_events",
      "required_inputs": [
        "user_id"
      ]
    },
    {
      "description": "Summarize tracked pipeline counts by stage for one user.",
      "name": "get_job_pipeline_summary",
      "required_inputs": [
        "user_id"
      ]
    },
    {
      "description": "Delete one cached search session or all sessions for a user.",
      "name": "clear_search_session",
      "required_inputs": [
        "user_id"
      ]
    },
    {
      "description": "Export all local records for a user across stores.",
      "name": "export_user_data",
      "required_inputs": [
        "user_id"
      ]
    },
    {
      "description": "Permanently delete all local records for a user.",
      "name": "delete_user_data",
      "required_inputs": [
        "user_id",
        "confirm"
      ]
    },
    {
      "description": "Suggest best outreach channel/contact for a job.",
      "name": "get_best_contact_strategy",
      "required_inputs": [
        "user_id"
      ]
    },
    {
      "description": "Generate a practical outreach draft tailored to user and role.",
      "name": "generate_outreach_message",
      "required_inputs": [
        "user_id"
      ]
    },
    {
      "description": "Start a background job search without requiring visa preferences.",
      "name": "start_job_search",
      "required_inputs": [
        "location",
        "job_title",
        "user_id"
      ]
    },
    {
      "description": "Poll incremental progress/events for a background job search run.",
      "name": "get_job_search_status",
      "required_inputs": [
        "user_id",
        "run_id"
      ]
    },
    {
      "description": "Fetch current result page from a background job search run.",
      "name": "get_job_search_results",
      "required_inputs": [
        "user_id",
        "run_id"
      ]
    },
    {
      "description": "Request cancellation of an in-progress background job search run.",
      "name": "cancel_job_search",
      "required_inputs": [
        "user_id",
        "run_id"
      ]
    },
    {
      "description": "Start a background search run for long scans.",
      "name": "start_visa_job_search",
      "required_inputs": [
        "location",
        "job_title",
        "user_id"
      ]
    },
    {
      "description": "Poll incremental progress/events for a background search run.",
      "name": "get_visa_job_search_status",
      "required_inputs": [
        "user_id",
        "run_id"
      ]
    },
    {
      "description": "Fetch current result page from a background search run.",
      "name": "get_visa_job_search_results",
      "required_inputs": [
        "user_id",
        "run_id"
      ]
    },
    {
      "description": "Request cancellation of an in-progress background run.",
      "name": "cancel_visa_job_search",
      "required_inputs": [
        "user_id",
        "run_id"
      ]
    },
    {
      "description": "Discover latest DOL LCA/PERM disclosure sources.",
      "name": "discover_latest_dol_disclosure_urls",
      "required_inputs": []
    },
    {
      "description": "Run internal pipeline to refresh sponsor-company dataset.",
      "name": "run_internal_dol_pipeline",
      "required_inputs": []
    },
    {
      "description": "Clear and reload in-memory company dataset cache.",
      "name": "refresh_company_dataset_cache",
      "required_inputs": []
    }
  ],
  "version": "0.3.1"
}

使用以下命令重新生成此部分和网站合同块：

python3 scripts/generate_contract_docs.py

验证生成的块是否为最新：

python3 scripts/generate_contract_docs.py --check

手动命令行界面（可选）

检查二进制版本：

visa-jobs-mcp --version

直接运行MCP服务器（用于调试）：

visa-jobs-mcp

运行内部DOL管道（维护人员工作流程，从源代码检查开始）：

./scripts/run_internal_pipeline.sh

故障排除

如果搜索未返回任何工作，请继续轮询 get_visa_job_search_status 并致电 get_visa_job_search_results 再次为相同 run_id.
如果发生上游速率限制，请等待几分钟，然后重试。
如果Homebrew安装因缺少发布资源而失败，请在发布工作流完成后重试。

数据与隐私

默认情况下，数据存储在本地。
不销售遥测或外部数据。
赞助匹配用途 data/companies.csv 以及基于DOL的流水线输出。

对于维护人员

Homebrew水龙头存储库： https://github.com/neosh11/homebrew-visa-jobs-mcp
贡献者指南： AGENTS.md
发布工作流程： .github/workflows/build-release-binaries.yml
本地发布工件构建： ./scripts/build_release_binaries.sh （原生Go二进制+捆绑 data/companies.csv)

许可证

MIT。看 LICENSE.

Visa Jobs MCP

快速接入

详细介绍