Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add yanhua1010/self-media-content-workflow --skill self-media-trend-radargit clone --depth 1 https://github.com/yanhua1010/self-media-content-workflowWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/yanhua1010/self-media-content-workflow/self-media-trend-radar)<a href="https://agentmods.dev/skills/yanhua1010/self-media-content-workflow/self-media-trend-radar"><img src="https://agentmods.dev/badge/skills/yanhua1010/self-media-content-workflow/self-media-trend-radar/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/yanhua1010/self-media-content-workflow/self-media-trend-radar"><img src="https://agentmods.dev/badge/skills/yanhua1010/self-media-content-workflow/self-media-trend-radar.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00125 | $0.00922 |
| Opus 5.5 | $0.00050 | $0.00369 |
| Sonnet 5.5 | $0.00025 | $0.00184 |
| Haiku 4.5 | $0.00013 | $0.00092 |
Grade A, and why
self-media-trend-radar scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
热点与竞品雷达
目标
识别用户需求、表达结构和时效机会，不复制竞品观点或文案。
数据源优先级
- 用户提供的链接、截图、导出文件和素材。
- 官方文档、原始发布、论文、公告和作者原帖。
- 平台公开搜索、热榜、竞品主页和公开互动数据。
- 可信二手资料，用于补充背景，不替代一手事实。
近期产品、人物、价格、版本、新闻和平台规则必须联网核验。技术事实优先使用官方来源。
海外对照模式（可选）
用户想知道某个话题在海外和国内的冷热差，或想找「海外已经热、国内还没人做」的选题时，且环境变量 AISA_API_KEY 已设置，可运行 scripts/overseas_compare.py，一次对比 X、YouTube、Reddit 与小红书、抖音、B站的近期热度。先用 quote 报价并取得用户同意，再执行 run。用法、费用、评分口径和限制见 overseas-compare.md。
没有 key 时不运行脚本，按下面的研究流程人工抽样。脚本结果作为第 2 步的样本，后续步骤不变。
Reddit 讨论发现（可选）
用户想在 Reddit 上宣传产品或内容时，先找出正在讨论相关问题的子版块和帖子，读评论和版规，给出可参与的清单，再交给平台文案写回复。流程、AIsa 接口和版规判断见 reddit-discovery.md。
研究流程
1. 定义研究问题
确认要回答的是热点窗口、关键词需求、竞品结构、账号定位还是内容机会。先限定平台、时间范围和样本数量。
2. 收集有限样本
只收集能回答问题的样本。记录链接、发布时间、标题、内容形态、公开互动和观察备注。不要把一次高表现直接叫作规律。
3. 拆解结构
逐条检查：
- 标题使用了什么承诺、身份、冲突或信息差。
- 开头 3 秒或第一屏如何建立停留。
- 内容如何组织证据、故事、步骤和转折。
- 封面或首图承担什么任务。
- 行动、标签、合集和发布时间如何配合目标。
- 评论区暴露了哪些真实问题和反对意见。
4. 判断机会
区分：
- 可以直接核验的事实。
- 多个样本重复出现的模式。
- 可能由热点、投流或账号体量造成的现象。
- 用户能加入的一手经验、反证和独立判断。
5. 转成原创选题
每个候选写明：一句话判断、目标受众、时效窗口、一手证据、差异化角度、适合平台、仍缺证据和风险。
输出
交付：
- 研究范围和样本说明。
- 3 到 5 条结构性发现。
- 不能下结论的内容。
- 3 到 5 个原创选题。
- 最推荐选题和下一步验证动作。
读取 research-safety.md 并遵守账号隔离、只读、控频、遇阻即停和凭证保护规则。
What ships with it
14 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- agents/openai.yaml 268 B
- references/overseas-compare.md 4.1 KB
- references/reddit-discovery.md 2.5 KB
- references/research-safety.md 1.2 KB
- scripts/overseas_compare.py 299 B runs code
- scripts/radar/__init__.py 59 B runs code
- scripts/radar/__main__.py 4.9 KB runs code
- scripts/radar/client.py 2.9 KB runs code
- scripts/radar/report.py 1.8 KB runs code
- scripts/radar/score.py 2.5 KB runs code
- scripts/radar/sources.py 7.2 KB runs code
- scripts/tests/__init__.py 0 B runs code
- scripts/tests/test_radar.py 3.7 KB runs code
- scripts/topics.example.toml 604 B
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 3d ago Changed · +10 lines · +29 tokens per session 7a6b0a865131
- 2mo ago First seen · 66 lines · 96 tokens per session scan A 23f3f1ea78c3
self-media-trend-radar is a skill published in the GitHub repository yanhua1010/self-media-content-workflow (647 stars, last pushed 3d ago), licensed MIT. It adds 125 tokens to every session and 922 once invoked, about $0.0005 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
self-media-trend-radar compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| self-media-trend-radaryanhua1010/self-media-content-workflow | Skill | 647 | 3d ago | MIT | A | 125 | Claude Code, Codex |
| instagram-post-builderSkillMedev/skills | Skill | 24 | 24d ago | MIT | A | 165 | Claude Code, Codex |
| instagram-warmupgenfeedai/skills | Skill | 3 | 2d ago | MIT | A | 80 | Claude Code |
| linkedin-warmupgenfeedai/skills | Skill | 3 | 2d ago | MIT | A | 81 | Claude Code |
| x-warmupgenfeedai/skills | Skill | 3 | 2d ago | MIT | A | 93 | Claude Code |
| youtube-warmupgenfeedai/skills | Skill | 3 | 2d ago | MIT | A | 77 | Claude Code |
| blog-content-creatorgenfeedai/skills | Skill | 3 | 2d ago | MIT | A | 55 | Claude Code |