autoxeo-agent
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@autoxeo-agentconnect my AutoXEO account and bind the current workspace to project 'seo-alpha'"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
AutoXEO Agent for Codex
A public Codex Plugin for running reproducible China GEO workflows. It includes six Codex-native Skills, a local stdio MCP, a companion workbench visible only on the loopback address, the Brand Wiki and Artifact protocols, and an AutoXEO Cloud collection adapter.
Issue production, Brand Wiki, analysis, retesting, and deliverables are all handled by the current Codex session; the plugin does not call a second conversation model, nor does it require DEEPSEEK_CHAT_API_KEY. Cloud is only responsible for account/project context, official platform API collection, Evidence, Job, and Credit Receipt.
The first real Provider slice is the DeepSeek official API. Doubao, Qwen, Tencent Yuanbao, and Kimi will show real configuration_required / eligibility_required statuses and will not pretend to be available until the official API and account are verified.
Local installation from GitHub Release
Download autoxeo-codex-plugin-marketplace-<version>.tar.gz and SHA256SUMS from the GitHub Release, verify them first, and then extract:
shasum -a 256 -c SHA256SUMS
tar -xzf autoxeo-codex-plugin-marketplace-<version>.tar.gz
codex plugin marketplace add /absolute/path/to/autoxeo-codex-plugin-<version>
codex plugin add autoxeo-agent@autoxeoAfter installing or upgrading, create a new Codex task. The distribution package only connects to https://agent.autoxeo.com by default. Do not write Provider Keys into the Plugin environment, Skills, or project reports; real collection credentials are configured only in AutoXEO Cloud.
Related MCP server: Codex MCP Server
Development environment integration
No mock collection is provided at runtime. When not connected to Cloud, the local Workspace, Brand Wiki, issue/analysis templates, and artifact preview remain available; collection, Credit, and authoritative status fail closed. Override the production default origin only when developing Cloud locally:
export AUTOXEO_CLOUD_BASE_URL=http://127.0.0.1:3000
# 可选;未设置时使用 ~/Documents/AutoXEO_Workspace
export AUTOXEO_WORKSPACE_ROOT=/absolute/path/to/AutoXEO_WorkspaceThen connect an existing AutoXEO account from the official website via start_account_connection or the WebUI Context page. Do not write credentials other than the Cloud origin into the environment, repository, Skills, reports, or diagnostic logs.
After logging in, first use get_account_overview to read the visible organizations and projects, then call bind_cloud_project (or select it in the WebUI's Context) to bind the local Workspace to a Cloud project, and explicitly set this task's Credit limit. If not bound, the issue set is not frozen, the balance is insufficient, or the platform capability is unavailable, production collection fails closed and no Credit is reserved.
Development and verification
npm ci
npm run check
npm run validate
npm run smoke:mcp
npm run smoke:webuiLocal preview:
npm run preview:webuiCalling open_local_workbench from Codex also starts the same WebUI; the MCP and the interface share project state and domain commands.
Repository and release boundaries
This repository is public and licensed under Apache-2.0; the AutoXEO name and trademark are not licensed with the code.
Cloud and Admin are separate private repositories and separate releases.
There is no fake collection at runtime; fixtures are only used for automated tests and cannot generate
observedEvidence.Payment and the four platforms pending integration must be shown as not yet configured and cannot be advertised as live.
Please report security issues privately via SECURITY.md; do not include Tokens, raw Evidence, customer files, or diagnostic logs in public Issues.
This server cannot be deployed
Maintenance
Related MCP Connectors
Build and supervise fleets of agents from Claude Code, Codex or Cursor. Connects over OAuth.
Discover, preview, estimate, run, and retrieve reusable AI workflows.
Search, read, and automate TextMine documents, records, workflows, integrations, and agent tasks.
- AurentiaOAuthfr.aurentia
Your Aurentia workspace — projects, CRM, tasks, deliverables — in Claude, Cursor or any MCP client.
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceConnects QGIS to Claude AI through the Model Context Protocol, enabling AI-assisted project creation, layer manipulation, processing algorithm execution, and Python code running within QGIS.7-
- AlicenseCqualityFmaintenanceConnects AI assistants like Claude to the Codex CLI for code analysis, editing, and execution. Supports file references with @ syntax, sandboxed code execution with approval workflows, and structured code changes for automated refactoring and documentation.8172 npm179MIT
- FlicenseNot gradedqualityDmaintenanceConnects OpenAI Codex Desktop to QGIS Desktop for natural language GIS operations, including project management, layer manipulation, and algorithm execution.4-
- AlicenseCqualityCmaintenanceEnables Claude and Claude Code to manage AACWorkflow tasks, agents, projects, squads, autopilots, and analytics through natural language.69MIT