setup-content-studio — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited setup-content-studio (Agent Skill) and scored it 87/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 3 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 3 flagged
The text {match} tells the agent to skip the normal "ask the user first" gate. Used adversarially it removes the human-in-the-loop check before destructive or sensitive actions, turning a normally-gated agent into a fire-and-forget executor.
The text {match} tells the agent to skip the normal "ask the user first" gate. Used adversarially it removes the human-in-the-loop check before destructive or sensitive actions, turning a normally-gated agent into a fire-and-forget executor.
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
You are helping set up a thought leadership content studio for a new person, using the content-studio plugin as the template.
This is an interactive, multi-step process:
Ask the user for:
Ask the user to provide example content. This is critical for voice/style analysis:
"Please share 15-30 example posts from this person. You can provide them as: - A text dump of their LinkedIn posts (with engagement numbers if available) - URLs to their LinkedIn posts - A document with collected posts
>
The more examples, the better I can capture their voice. Include engagement metrics (reactions, comments, reposts) if you have them - this helps identify what resonates with their audience."
Do not proceed until you have at least 10 substantive example posts. Shares/reposts with minimal commentary don't count. Unless the person does not have at least 10 posts.
Use WebSearch to find:
Combine the online research with what you learn from their example posts.
Read references/voice-analysis.md for the full analysis framework covering writing style, content themes, and engagement patterns.
Based on your analysis, ask the user questions to fill gaps. Use the AskUserQuestion tool with relevant options. Key questions to consider:
If not clear from examples, ask:
"What types of content should the studio support?" Options: LinkedIn posts only / LinkedIn + blog posts / LinkedIn + blog + opinion pieces / Other
If not clear from examples:
"What language(s) should the content be in?" Options: English only / English + Dutch / English + [other] / Multiple
If opinion pieces are a content type, also ask which language the opinion pieces should be written in (it may differ from LinkedIn post language). Use this to set {{OPINION_LANGUAGE}} in guidelines/opinion.md.
If the examples show a consistent pattern, confirm it. If not, ask:
"Their posts range from X to Y words. What's the ideal target range?"
"Besides LinkedIn, where does this person publish?" (Blog, newspaper columns, industry publications, etc.)
Only ask questions where the answer isn't already clear from the example posts and research. Don't ask about things you can confidently infer.
Read references/repo-setup.md for the full repository setup procedure, including template copying, file personalization, skill creation, and hook configuration.
Convert each example post into a YAML file. Read references/content-format.md for the file path convention and YAML structure.
If your agent supports sub-agents (e.g. Claude Code), use parallel Task agents to create posts in batches of 10 for efficiency. Otherwise (e.g. Codex), create the posts sequentially: the output is identical, it just takes longer.
cd content-studio
npm installTest the scripts work:
./scripts/search-posts.sh "keyword"
./scripts/list-published.sh linkedin-postgit add .
git commit -m "Set up content studio for [Person Name]
Adapted from content-studio plugin with:
- Professional profile and writing guidelines for [Name]'s voice
- [N] published LinkedIn posts with full text and engagement metrics
- Skills for [list skills]
- Content types: [list types]
Co-Authored-By: Claude Opus 4.6 <[email protected]>"
git push -u origin mainPresent a summary to the user:
cd content-studio && npm run dev)/write-linkedin-post, etc.)~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.