Home / All Skills / Content & Media Creation / dbs-content
English (EN)简体中文 (ZH)日本語 (JA)
Content & Media Creation · dontbesilent2025/dbskill

dbs-content

Diagnoses how to turn a determined topic into quality content. Use when users ask to design content, review copywriting, or improve content delivery and expression.

L1 Basic · Source registration onlyV1 Commit-PinnedDual-pass Audit
Author dontbesilent2025GitHub dontbesilent2025/dbskill ↗Stars 10121License CC BY-NC 4.0（仓库根 LICENSE 首行「Attribution-NonCommercial 4.0 International (CC BY-NC 4.0)」，条款为 Attribution + NonCommercial，Copyright (c) 2026 dontbesilent。**非商业授权**：个人/学习/研究可直接使用，商业用途需单独授权；GitHub API license.spdx_id 为 NOASSERTION（CC 系非 SPDX 标准标识））Commit 8b8e33f1ec
Control levels describe skill capabilities and suggested control measures; they do not represent quality or safety certifications, nor do they dictate usability. Usage decisions rest entirely with users and administrators. Different agent hosts apply varying constraints; do not assume hosts provide automatic protection. Third-party skill copyrights belong to their original authors. Use at your own risk.

1How It Works & Mechanism

Positioned as a content diagnostician rather than a ghostwriter: only evaluates direction, format, and expression, leaving the actual writing to the user.

skills/dbs-content/SKILL.md
**你不帮人写内容。你帮人诊断内容该怎么做。**
What it does: breaks down 'is the content good' into three diagnosable layers—direction (is the topic worth pursuing?), format (what medium and platform should be used?), and expression (textual cleanliness, cover/title, delivery efficiency, and cognitive gap)—delivering actionable evaluations without drafting. The core prohibition 'Never write content for the user—you are a diagnostician, not a ghostwriter' reinforces this positioning.

Prerequisite gatekeeping: assumes the user already has a clear topic; clarifies it if missing, or skips directly to the 5-dimension diagnosis if a draft exists.

skills/dbs-content/SKILL.md
**前提：用户应该已经有一个明确的选题。**
Phase 1 branches into three paths based on input: no topic → clarify the general direction first; topic selected without format → match format first (Phase 2); draft exists → jump directly to Phase 3 for diagnosis ('If the user has a first draft → jump directly to Phase 3 for diagnosis.'). Triage prevents forcing a report under insufficient information.

Phase 2 features two matching tables: 'Topic Attributes → Content Format' and 'Platform Attributes → Suitable Format', detailing rationales for every row.

skills/dbs-content/SKILL.md
| 观点输出、认知冲突 | 短视频（露脸口播） | 人的表情和语气是最强的说服工具 |
The format table has 5 rows: Opinion Pieces, Tool Lists, Deep Dives, Case Studies, and Controversial Topics, each paired with recommended formats and a one-line rationale. The platform section defines positioning (e.g., 'Xiaohongshu = content platform, best for graphic posts and short videos, follower growth relies on content quality'; 'Douyin = content e-commerce platform, best for short videos and live streams, monetization relies on paid ads'). This transforms 'what format to use' from an aesthetic debate into a structured table lookup.

Phase 3: 5-dimension diagnosis covering Text Cleanliness → Cover/Title → Delivery Efficiency → Cognitive Gap → AI-assisted Creation Suggestions, providing tier ratings and self-check questions for each.

skills/dbs-content/SKILL.md
#### 维度 4：认知落差检测
Dimension 1 checks for AI tone and the 'actionable-content trap', citing Wittgenstein (language must be publicly verifiable); Dimension 2 checks if the cover is image-led or text-led, and if the title informs or hijacks cognition; Dimension 3 asks 'Can the core point be stated in one sentence?' and 'Are you spending 99% of your time wrapping 1% of content?'; Dimension 4 checks if peers have already covered it thoroughly and whether audiences will feel 'I already know this'; Dimension 5 assigns distinct AI workflows based on opinion, analysis, tool, or tone-optimization types.

Phase 4 outputs a standardized diagnosis report: Recommended Format + 5-Dimension Evaluation Table + 'What is the first step?' + a one-sentence summary.

skills/dbs-content/SKILL.md
| 文字洁癖 | ✅/⚠️/❌ | {具体问题} |
The report template embeds a code block (# Content Creation Diagnosis Report: {Topic Name}), where '## If you are going to make this, what is the first step?' strictly demands 'a concrete action, not an advisory suggestion'—turning the report from critique into actionable items.

Special warning list converts 5 types of high-risk inquiries into standardized pushback scripts, preventing users from derailing diagnosis into title fixation, generic 'value-drop' traps, or blaming AI shadowbans.

skills/dbs-content/SKILL.md
用户纠结标题怎么写 → **「99% 时间应该花在做值得讲的事上，标题只需要 1%。你在纠结的是那 1%。」**
The other four are: wanting to share generic 'value/actionable tips' → reject ('All creators telling you to give pure value are unprofessional'); claiming 'AI content got shadowbanned' → attribute to the author lacking textual cleanliness; attempting content without a product → product comes before content ('Where is your payment link?'); aiming for emotional/personal growth topics → warn that while topic volume is sufficient, value validation is extremely hard. These serve as gateways to the explanatory framework, not generic advice.

Inlined case library (3 positive, 2 negative cases with diagnostic takeaways), serving to validate principles and provide tone exemplars.

skills/dbs-content/SKILL.md
> 让 AI 分析爆款文案 = 最蠢方法。规则 > 结果。
Positive cases include 'Revised cover title, tripled views', 'Xiaohongshu account gained 3,200 followers in 48h', and 'Sold first AI agent within 24h'; negative cases include the death spiral and the aforementioned 'Asking AI to analyze viral copy'. Diagnosis points cross-reference Principles 2/3/4/1, building a bidirectional 'principle-to-case' index.

2Core Capabilities

01Topic → content format matching (recommended formats and rationales for 5 topic types)
02Platform compatibility assessment (attributes and suitable formats for Xiaohongshu, Douyin, X, and WeChat Official Accounts)
035-dimension content diagnosis (Text Cleanliness, Cover/Title, Delivery Efficiency, Cognitive Gap, AI Workflow), each rated with ✅/⚠️/❌
04Standardized diagnosis report template (including concrete action for 'What is the first step?')
055 special warning scripts (title fixation, generic value trap, shadowban blame, content without product, emotional/personal growth)
06Inlined positive and negative case library (including the death spiral and 'analyzing viral copy with AI' negative cases)

4Recommended Governance & Evidence

Suggested Controls for Safe Operation
For Agent Runtime
Can be auto-loaded
For Administrators
Register source and version
Evidence & Triggers (File Paths & Excerpts)
  • SKILL.md: Pure prompt instructions and rule text, without script execution or external impact
Audit Notes: Evaluated under Batch 4 Final Tier 1 criteria (pure prompt/instruction: no outbound traffic, no disk writes, no credentials, no reversing/spoofing/bypassing): This skill directory contains only SKILL.md and agents/openai.yaml, with no scripts, references, templates, or data files. Runtime triggers no tools, network calls, credential reading, or disk writes. Tier exception verification: SKILL.md suggests a user-built workflow only under 'Dimension 5: AI-assisted Creation Suggestions' (source text: 'get notes to extract material → deep search with AI → agent rewrites into short video script → face-to-camera shooting'), which advises the user's content production pipeline. It neither mandates a generative backend called by this skill at runtime, nor contains URLs/interfaces, nor transmits user material to external services. It does not meet orange criteria (designated generative backend + outbound material + quota consumption), thus remaining green. Residual risks are limited to prompt-level logic and content stances (see risks).

5Second-Pass Verification

  • [ok] 无脚本/无网络/无凭证/无写入（green 档自证） — 目录仅 SKILL.md + agents/openai.yaml；token 扫描（http|curl|wget|API_KEY|token|secret|\.env|bash|exec|python|~/\.）零命中；SKILL.md 无工具调用与读写文件指令。
  • [ok] 「AI 辅助创作建议」是否触发 orange（核查） — 该节是给用户的内容生产流程建议（get 笔记 / 各家 AI deep search / 智能体改写），无 URL、无 API 名、无「调用」指令，也不涉把用户素材送出；不满足 orange 要件，维持 green，已在 reason 写明核查过程。
  • [ok] 五维诊断确为 5 维且各有判定档 — 文字洁癖 / 封面标题 / 表达效率 / 认知落差 / AI 辅助创作建议，五节逐段读过；前四维给出 ✅/⚠️/❌ 判定口径，第五维给工作流建议。
  • [ok] 报告模板与警告话术逐字存在 — Phase 4 代码块含推荐形式、五维表、「第一步是什么」与「一句话」；特别警告 5 条含固定台词，均逐字核对。
  • [ok] 案例库内容为真实节选而非编造 — 正例 3 / 反例 2 均逐字出现在「内联案例库」小节，含数据细节（点击率 5%→19%、48h 3200 粉等），未发现虚构案例。
  • [ok] 与 marketplace 条目描述一致 — marketplace.json 描述「dontbesilent 内容创作诊断。选题通过后，诊断怎么把这个选题做成好内容。」与 SKILL.md「前提：用户应该已经有一个明确的选题」一致。
  • [ok] 元数据（license/stars/last_push/installs/commit） — GitHub API：stars 10121、pushed_at 2026-09-07T03:48:59Z、license.spdx_id NOASSERTION（根 LICENSE 为 CC BY-NC 4.0，非商业）；本地 HEAD = 8b8e33f1ecaed8cee606fe950c4426b525ead314 = 任务 pin。
  • [discrepancy] meta.installs 值域校正（批4 质检 2026-09-15） — 原记录为 skills.sh 仓库级聚合徽标（579.6K（skills.sh 徽标 /b/dontbesilent2025/…），并非本 skill 的单体安装量；站内惯例只记可定位的单 skill 数值 → 校正为 null（其他 25 份 dbs 报告本就为 null，修此 8 份以齐平）

6Verdict & Scenarios

  • Diagnoses without ghostwriting; deliverable is clear evaluation + first-step action, preventing unusable ghostwritten drafts.
  • Tabular matching for formats and platforms, reducing circular discussions on 'which format to pick'.
  • 5-dimension diagnosis breaks down 'is content good' into 4 self-check dimensions + 1 AI workflow recommendation, anchoring evaluations.
  • Prerequisite gatekeeping and triage prevent forcing reports when information is insufficient (clarifies missing topics, diagnoses drafts directly).
  • Reports output actionable items: concrete first action + one-sentence summary for immediate execution.
    • Best for: Ideal for content creators (WeChat Official Accounts, short video, Xiaohongshu, Twitter/X) with a determined topic looking for a direction and format health-check before production, or troubleshooting 'what feels off' in a first draft; especially helpful for writers endlessly tweaking titles without writing content—pulling them back to topic and substance via the '99% vs. 1%' principle. Also fits content teams needing a consistent diagnostic framework (format table + 5-dimension table + action items) for self-inspection checklists. A solid reference model for authors learning how diagnostic skills structure subjective evaluations using tables.
    Not for: Not suitable for stages where no product exists and content is expected to bootstrap traction upfront (SKILL.md explicitly demands 'product before content' and redirects you to build the product first); not suitable for users seeking ghostwriting (strictly does not draft content); not suitable for pre-publishing checks regarding sensitive terms, advertising law, or platform violations (use dbs-content-risk-check instead); not suitable for content decisions requiring data backing (no competitor checks or data validation); and not intended for treating 'cognitive gap' assessments as factual conclusions.
    Install Agent-Ready
    ① Verified Mirror 2026-09-15
    Option A · Direct DownloadDownload dbs-content.tar.gz
    sha256: 4edb918184680335…
    Option B · Copy Guide for Agent
    install.json
    Agents automatically download and verify mirrors from this JSON guide.
    ② Upstream GitHub
    GitHub Repo ↗
    Pinned to commit 8b8e33f1ec.