菜单
Research and draft approval-gated social and blog content for the seeded GTM campaign using current public evidence.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/diggerhq/opencomputer-example-gtm --skill gtm-content命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Safely combine optional Composio connected accounts and bounded Apify harvests with the GTM workflows, including provider availability, read/write boundaries, and explicit opt-in configuration.
Search and research the current web with Exa using ranked sources, semantic queries, domain and publication-date filters, category targeting, and extracted highlights or full text. Use for current facts, news, source discovery, literature searches, company or people research, comparisons, and any request that needs live web evidence.
Run the dry-run social opportunity workflow for the seeded agent-harness campaign, including scheduled discovery, ranked shortlists, human selection, and response drafting.
正在显示 SKILL.md
| name | gtm-content |
| description | Research and draft approval-gated social and blog content for the seeded GTM campaign using current public evidence. |
For create_social_content, research current evidence and propose no more than
five concepts. Each concept must include its intended audience, destination,
source URL, concrete angle, and timeliness rationale.
After the human selects a concept, draft only that concept. Keep X and LinkedIn variants distinct. Never claim publication or scheduling.
For research_blog_topics, propose no more than five topics. Each needs:
Do not claim access to Search Console, keyword volume, private analytics, or private repository work unless the current input actually contains that data. Selection authorizes a detailed outline and opening section, not publication.