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 TestAny-io/testany-agent-skills --skill media-writergit clone --depth 1 https://github.com/TestAny-io/testany-agent-skillsWrote 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/testany-io/testany-agent-skills/media-writer)<a href="https://agentmods.dev/skills/testany-io/testany-agent-skills/media-writer"><img src="https://agentmods.dev/badge/skills/testany-io/testany-agent-skills/media-writer/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/testany-io/testany-agent-skills/media-writer"><img src="https://agentmods.dev/badge/skills/testany-io/testany-agent-skills/media-writer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00050 | $0.01515 |
| Opus 5.5 | $0.00020 | $0.00606 |
| Sonnet 5.5 | $0.00010 | $0.00303 |
| Haiku 4.5 | $0.00005 | $0.00152 |
Grade A, and why
media-writer 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 1mo 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.
How it starts
The opening of the file, as written. The whole thing — 72 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Media Writer - 自媒体写作工作流
作为内容生产的协调者，按用户请求完成必要创作、事实核验与编辑，不把八阶段目录当作每次必经的审批链。
先确定执行模式
执行前读取 写作执行与授权约定，所有阶段均遵循它。
| 模式 | 使用场景 | 完成条件 |
|---|---|---|
end_to_end |
写完文章、连续完成必要编辑 | 交付成稿或明确事实缺口的草稿，不逐阶段泛问 |
checkpointed |
逐阶段确认或指定检查点 | 在用户指定检查点停下，不提前写后续产物 |
single_stage |
只做选题、Brief、角度分析、审校等 | 完成该项即结束，不扩展任务 |
已明确的模式、主题、平台、立场、材料及输出路径直接复用。对具体提案回复“可以”按上下文识别授权，不索要固定口令；不把称赞或沉默解释为扩大范围。先读已有材料，只问确有影响的缺口。
阶段与资源
按实际需要读取对应 prompt，不能凭角色名称假装已执行。路径相对本次安装的 SKILL 文件，不相对产品仓库。
| Stage | 职责 | 输入 | 产物 | Prompt |
|---|---|---|---|---|
| 1 | Topic Scout | 用户想法或既有 Brief | 明确主题、受众、立场、素材需求 | references/prompts/01-topic-scout.md |
| 2 | Researcher | Brief、给定素材 | 可追溯事实、争议与缺口 | references/prompts/02-researcher.md |
| 3 | Strategist | Brief + 素材 | 写作角度、大纲和平台策略 | references/prompts/03-strategist.md |
| 4 | Writers | 已选角度、平台、素材 | 平台草稿 | references/prompts/04-writer-{platform}.md |
| 5 | Selector | 多篇候选草稿 | 选择理由与候选稿 | references/prompts/05-selector.md |
| 6 | Editors | 候选稿或给定原稿 | 逻辑、风格、细节检查后的稿件 | references/prompts/06-logic-editor.md、06-style-editor.md、06-detail-editor.md（同目录） |
| 7 | Illustrator | 终稿及明确配图请求 | 图片来源或配图 prompt | references/prompts/07-illustrator.md |
| 8 | Archivist | 明确归档请求与项目文件清单 | 归档及验证记录 | references/prompts/08-archivist.md |
Stage 1-3 的输入已齐备可直接复用，不强制重选题或重新询问平台。只有一篇已选草稿不需要人为生成多稿再筛选。Stage 6 在完整成稿时按逻辑→风格→细节检查，不等于三次审批。用户只要求单项审校时仅做所请求维度。
Stage 7/8 仅在请求包含时适用。纯文本文章不要求配图或归档，更不执行发布、发送或付费生成。归档不得移动无关文件；具体边界见执行约定与手册。
调度与工具回退
- 多平台分别生成适配稿件；有实际且获准的独立 agent 能力时可并行，否则顺序完成。不得因没有
Task而停止，也不得虚构并行、独立审查或固定倍数加速。 - 有委派时传递模式、范围、检查点、输入和输出路径；子 agent 只返回分配的阶段，协调者按模式继续，不继承“每阶段必须等用户”的旧停顿。
- 文件读取、提问、任务追踪和检索采用当前真实可用能力。
Read、TodoWrite、AskUserQuestion等只是名称示例。
平台与作者材料
写作前读取目标平台 references/platforms/{platform}-guide.md 以及作者风格材料：
references/persona/my-voice.mdreferences/persona/my-values.mdreferences/persona/my-audience.md
平台标识：wechat、zhihu、xiaohongshu、linkedin、medium、reddit。作者人设和示例只指导风格，不证明作者有某种经历、测量结果或业绩；具体事实以获准材料为准。用户明确的字数、受众和立场优先于通用长文建议。
What ships with it
46 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 288 B
- assets/angle-analysis-template.md 10 KB
- assets/brief-template.md 8.5 KB
- assets/draft-checklist.md 10 KB
- assets/testany-logo-small.png 48 KB
- assets/testany-logo.svg 7.5 KB
- references/execution-modes.md 4.4 KB
- references/orchestrator-manual.md 8.4 KB
- references/persona/my-audience.md 7.4 KB
- references/persona/my-values.md 6.3 KB
- references/persona/my-voice.md 6.4 KB
- references/persona/past-articles/README.md 1.9 KB
- references/platforms/linkedin-guide.md 16 KB
- references/platforms/medium-guide.md 17 KB
- references/platforms/reddit-guide.md 17 KB
- references/platforms/wechat-guide.md 12 KB
- references/platforms/xiaohongshu-guide.md 12 KB
- references/platforms/zhihu-guide.md 14 KB
- references/prompts/01-topic-scout.md 25 KB
- references/prompts/02-researcher.md 17 KB
- references/prompts/03-strategist.md 14 KB
- references/prompts/04-writer-linkedin.md 17 KB
- references/prompts/04-writer-medium.md 17 KB
- references/prompts/04-writer-reddit.md 14 KB
- references/prompts/04-writer-wechat.md 17 KB
- references/prompts/04-writer-xiaohongshu.md 13 KB
- references/prompts/04-writer-zhihu.md 19 KB
- references/prompts/05-selector.md 13 KB
- references/prompts/06-detail-editor.md 14 KB
- references/prompts/06-logic-editor.md 14 KB
- references/prompts/06-style-editor.md 16 KB
- references/prompts/07-illustrator.md 10 KB
- references/prompts/08-archivist.md 10 KB
- scripts/check-platforms.sh 1.1 KB runs code
- scripts/check-prompts.sh 1.1 KB runs code
- scripts/check-structure.sh 504 B runs code
- scripts/check-workflow-status.sh 886 B runs code
- scripts/count-drafts.sh 582 B runs code
- scripts/create-dirs.sh 383 B runs code
- scripts/credentials.py 5.5 KB runs code
- scripts/image_download.py 4.5 KB runs code
- scripts/image_generate.py 10 KB runs code
- scripts/image_get.py 6.7 KB runs code
- scripts/image_search.py 9.2 KB runs code
- scripts/requirements.txt 317 B
- scripts/start-workflow.sh 537 B runs code
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.
- 1mo ago Changed · -211 lines · +10 tokens per session 18536f896576
- 1mo ago First seen · 283 lines · 40 tokens per session scan A e679d3eb3c91
media-writer is a skill published in the GitHub repository TestAny-io/testany-agent-skills (82 stars, last pushed 4d ago), licensed MIT. It adds 50 tokens to every session and 1,515 once invoked, about $0.0002 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 2026-09-08.
Other skills, from other repositories
web-test
A browser-automation tool for testing the 1C:Enterprise web client. It uses Playwright to navigate sections, fill forms, read tables and reports, and perform user actions in a browser.
form-edit
A tool for modifying an existing managed form in 1C by adding screen elements, fields, commands, and event handlers. Event handlers are code routines that run when something happens, such as a field changing.
cfe-patch-method
A generator for method interceptors in 1C extensions, which let an extension add code before or after an existing method or replace its body. It reads the original method and creates or updates the required BSL source file.
db-list
A command for listing and managing the 1C databases registered in a project's .v8-project.json file.
form-patterns
A reference guide for arranging managed forms in the 1C platform. It describes common layouts, naming conventions, events, properties, and advanced patterns for different form types.
mxl-compile
A generator for 1C MXL spreadsheet-document layouts from a compact JSON description. MXL is the 1C format used for tabular documents such as printable forms.