Install
openclaw skills install market-kit-skillsUse when the user needs marketing deliverables such as campaign plans, Xiaohongshu notes, audience positioning, selling-point refinement, reference-grounded copy, or marketing images.
openclaw skills install market-kit-skillsMarket Kit Skills 是一个面向营销场景的生产型 skill,适合在需要明确产出时使用,而不是停留在泛化讨论。
它最擅长这些事情:
Use the bundled scripts to inspect optional context, submit the task, and fetch the final result:
list_projects.py 查看可选资料库list_skills.py 查看可选技能chat.py 提交任务并拿到 conversation_idchat_result.py 查询最终结果scripts/list_projects.py,选择一个或多个 project_idscripts/list_skills.py,选择一个或多个 skill_idscripts/chat.py,传入 --message,必要时再传 --project-id、--skill-idconversation_idscripts/chat_result.py --conversation-id ... 获取结果conversation_idconfirm_info,可以继续发送自然语言修订,也可以通过 form_id + form_data 结构化续跑branch 表示实际走到的营销分支,比如 collect_info、confirm_info、generate_plan、generate_notes、generate_imageresult 是首选的结构化结果text 只作为兜底摘要conversation_id 必须保留,用于后续续聊web_url 是网页版结果链接,格式为 https://justailab.com/marketing?conversation_id=<conversation_id>result.result.components[].data.images[].urlresult.result.components[].data.titleresult.result.components[].data.contentList projects:
python3 "${CLAUDE_SKILL_DIR}/scripts/list_projects.py"
List skills:
python3 "${CLAUDE_SKILL_DIR}/scripts/list_skills.py"
Run a new turn:
python3 "${CLAUDE_SKILL_DIR}/scripts/chat.py" --message "帮我做一份新品 campaign plan"
Poll the result:
python3 "${CLAUDE_SKILL_DIR}/scripts/chat_result.py" \
--conversation-id "existing-conversation-id"
Continue an existing turn:
python3 "${CLAUDE_SKILL_DIR}/scripts/chat.py" \
--conversation-id "existing-conversation-id" \
--message "把内容方向改成更适合小红书种草的表达"
project_idskill_idresult,不要只读顶层 textchat_result.py 还在输出 status=running,说明营销内容仍在生成,不能过早判断“没有图片”或“没有结果”generate_notes、generate_image 这类慢分支,除非用户明确要求,否则不要把 chat_result.py --timeout 设成小于 300Polling timed out before task completed.,不要把轮询超时当成任务失败;这通常表示当前轮询窗口不够长,任务仍可能在后台继续生成running 或出现轮询超时时,应明确告诉用户“还在生成”,不要自己擅自生成标题、正文、图片说明或图片链接冒充最终结果web_url