platforms/codex/skills/mx-interaction/SKILL.md
Safely work with Midea MX / 美信 interactions from Codex. Use when the user asks to search MX users or groups, prepare or send an MX message, notify someone in MX, inspect MX message-query capabilities, search MX chat history, or understand MX ocToken / UA auth behavior. Prefer dry-run before any live send.
npx skillsauth add codingsamss/ai-dotfiles mx-interactionInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Use this skill for MX / 美信读写交互. Keep all token handling local and explicit.
scripts/mx_safe_send.py for message sends. It dry-runs by default.--send and --confirm-user-authorized.mx_message_* tools only.mx_message_* tools are available in the current Codex context, say that the current context has no MX query tools and stop.references/auth-and-token.md.references/message-query.md.scripts/user_search.py.scripts/group_search.py.python3 scripts/mx_safe_send.py \
--from-id "<sender_uid>" \
--to-id "<receiver_uid_or_group_id>" \
--session-type p2p \
--body "message text"
python3 scripts/mx_safe_send.py \
--from-id "<sender_uid>" \
--to-id "<receiver_uid_or_group_id>" \
--session-type p2p \
--body "message text" \
--send \
--confirm-user-authorized
For file, image, @user, @all, or push fields, read references/auth-and-token.md and use the copied original scripts only when the user explicitly asks for those behaviors. The original im_send.py also requires --confirm-user-authorized for any live send.
Search users:
python3 scripts/user_search.py --keyword "<姓名/工号/邮箱前缀>"
Search groups:
python3 scripts/group_search.py --keyword "<群名关键字>" --limit 5
These commands require a valid token from MX_OC_TOKEN, the current MX/OpenClaw runtime MCP config, or /tmp/ua_token.txt; do not run them if the user asked for static analysis only.
Use mx_message_sessions, mx_message_list, mx_message_search_global, mx_message_search_session, and mx_message_history only when those tools are exposed in the current Codex tool list.
Do not emulate those tools with filesystem reads. mx_team_info and mx_team_search are documented in the MX package but are not exported by the current dist/src/tools/index, so treat them as unconfirmed until live tool availability proves otherwise.
tools
Apifox 接口测试用例:test-case 与 test-data 的查询、创建、更新、删除、分类和运行;处理测试步骤、断言、提取变量、前后置处理器、数据集,以及“CLI 创建后前端测试步骤无法展示”等问题。
tools
通过 Apifox CLI 管理 Apifox 项目资源。触发场景:运行接口自动化测试/测试套件,查询/创建/更新/删除接口、环境、Schema、Mock、分支等项目资源,导入导出 API 文档,查看测试报告,管理 Runner、定时任务、通知等 CI/CD 配置。CLI 输出为结构化 JSON,常含 agentHints.nextSteps;所有命令支持 --help。
tools
Apifox 接口测试用例:test-case 与 test-data 的查询、创建、更新、删除、分类和运行;处理测试步骤、断言、提取变量、前后置处理器、数据集,以及“CLI 创建后前端测试步骤无法展示”等问题。
tools
通过 Apifox CLI 管理 Apifox 项目资源。触发场景:运行接口自动化测试/测试套件,查询/创建/更新/删除接口、环境、Schema、Mock、分支等项目资源,导入导出 API 文档,查看测试报告,管理 Runner、定时任务、通知等 CI/CD 配置。CLI 输出为结构化 JSON,常含 agentHints.nextSteps;所有命令支持 --help。