local-article-tool
Skill0 starsChanged 6 months agoMITView on GitHub
Local-first workflow for turning an X link, article link, or short brief into one full standalone article in OpenClaw voice.
In this repository15 other files it gives its agents.
- approved-backlog-watchdogSkill
- context-engineeringSkill
- elon-operatorSkill
- goal-modeSkill
- image-readingSkill
- mac-ttsSkill
- personal-knowledge-wikiSkill
- repurpose-funnelSkill
- skill-builderSkill
- swipefile-compilerSkill
- thumbnail-creatorSkill
- video-to-skillSkill
- visual-first-explainerSkill
- x-link-best-effortSkill
- youtube-learnSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
26 rules. About 0 are common, 24 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈24
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 6
Use this skill when Charlie wants a real article draft, not a thread.
- Distinctive: line 9
Default to one full standalone article.
- Distinctive: line 10
Do not output hooks, single-post versions, swipefile tags, or a thread unless Charlie explicitly asks for them.
- Very specific: line 11
When the source is an X link, call x_tweet_fetch first.
- Distinctive: line 12
If X fetch is partial or blocked, continue from the fetched evidence and conservative inference.
- Distinctive: line 22
Keep the OpenClaw identity and soul in tone, conviction, rhythm, and perspective.
- Distinctive: line 23
Use the identity as seasoning, not as a costume.
- Distinctive: line 25
Write from inside the climb, not from above it.
- Distinctive: line 27
When a prior article pattern proved strong, reuse its underlying moves, not its wording.
- Distinctive: line 28
When the source is factual news or a product/company change, keep factual integrity above style.
- Distinctive: line 35
Start with tension immediately.
- Distinctive: line 38
Keep the piece article-length by default.
- Distinctive: line 40
Avoid business-book cliches, polished abstractions, and tidy contrast lines that sound manufactured.
- Distinctive: line 41
If a line sounds like generic leadership advice instead of a real operator observation, rewrite it.
- Very specific: line 42
Before returning any full article, silently run the checks in references/x_article_preflight.md.
- Distinctive: line 43
If any check fails, revise before returning.
- Distinctive: line 44
If the draft sounds like AI self-help, LinkedIn coaching, a summary, or a slogan-heavy mini-rant, rewrite it before returning.
- Distinctive: line 51
Avoid bullet-list-heavy output unless the user explicitly wants that format.
- Distinctive: line 52
No CTA or follow-up question unless the user asks for one.
- Distinctive: line 67
Do not ask for angle selection on a bare source.
- Distinctive: line 68
If the source was already seen or drafted before, use that context silently and still return the article instead of offering alternatives.
- Distinctive: line 71
Use previous drafts silently as negative examples and make the new draft stronger in voice, specificity, texture, and closing.
- Distinctive: line 73
Do not let a factual article turn into a checklist, audit handout, or incomplete numbered playbook unless Charlie explicitly asked for that …
- Distinctive: line 74
Before returning, verify the article is complete: no truncated lists, no broken final sentence, and no raw markdown-style URLs in the prose.
- Distinctive: line 76
Prefer narrow, defensible wording over dramatic wording.
- Distinctive: line 77
Never imply ownership or acquisition unless the source explicitly supports it.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Use this skill when Charlie wants a real article draft, not a thread.
Primary behavior:
- Default to one full standalone article.
- Do not output hooks, single-post versions, swipefile tags, or a thread unless Charlie explicitly asks for them.
- When the source is an X link, call
x_tweet_fetchfirst. - If X fetch is partial or blocked, continue from the fetched evidence and conservative inference.
- Learn from the source, but do not summarize or rewrite it verbatim.
- Extract the thesis and rebuild it as a fresh article.
Voice:
- Read
/Users/carlosbuelvas/.openclaw/workspace/IDENTITY.md. - Read
/Users/carlosbuelvas/.openclaw/workspace/SOUL.md. - Read
references/audience_climbing_builder.md. - Read
references/high_signal_article_keynotes.md. - Read
references/fact_integrity_source_articles.md. - Keep the OpenClaw identity and soul in tone, conviction, rhythm, and perspective.
- Use the identity as seasoning, not as a costume.
- The result should read like a publishable article, not roleplay.
- Write from inside the climb, not from above it.
- Sound like one of the disciplined, ambitious people we are writing for, not like a guru or genius explaining life to them.
- When a prior article pattern proved strong, reuse its underlying moves, not its wording.
- When the source is factual news or a product/company change, keep factual integrity above style.
X article mode:
- Also read
references/x_article_attention_spec.md. - Also read
references/x_article_preflight.md. - For X-sourced or attention-first article tasks, optimize for attention capture and retention, not summary completeness.
- Make one sharp argument.
- Start with tension immediately.
- Escalate with proof, contrast, and implication.
- End with a hard closing line.
- Keep the piece article-length by default; do not collapse it into a short manifesto unless Charlie explicitly asks for short.
- Develop the thesis with example, contrast, implication, business stakes, or operator consequences before closing.
- Avoid business-book cliches, polished abstractions, and tidy contrast lines that sound manufactured.
- If a line sounds like generic leadership advice instead of a real operator observation, rewrite it.
- Before returning any full article, silently run the checks in
references/x_article_preflight.md. - If any check fails, revise before returning.
- If the draft sounds like AI self-help, LinkedIn coaching, a summary, or a slogan-heavy mini-rant, rewrite it before returning.
Quality bar:
- Strong opening hook.
- Clear thesis.
- Tight flow from insight to implication to action.
- No filler.
- Avoid bullet-list-heavy output unless the user explicitly wants that format.
- No CTA or follow-up question unless the user asks for one.
Internal references:
references/x_article_attention_spec.mdreferences/audience_climbing_builder.mdreferences/high_signal_article_keynotes.mdreferences/fact_integrity_source_articles.mdreferences/x_article_preflight.mdtemplates/article_template.mdtemplates/review_checklist.md
Default output:
titlefull article- Do not ask for angle selection on a bare source. Choose the strongest article angle yourself and draft immediately.
- If the source was already seen or drafted before, use that context silently and still return the article instead of offering alternatives.
- Repeated-source sends mean improve the article, not restart from zero and not offer menus.
- Use previous drafts silently as negative examples and make the new draft stronger in voice, specificity, texture, and closing.
- For factual/news-driven pieces, avoid sensational headline language unless the source clearly supports it.
- Do not let a factual article turn into a checklist, audit handout, or incomplete numbered playbook unless Charlie explicitly asked for that format.
- Before returning, verify the article is complete: no truncated lists, no broken final sentence, and no raw markdown-style URLs in the prose.
- In factual/news-driven pieces, mark the difference between what the source reported and what we infer from it.
- Prefer narrow, defensible wording over dramatic wording.
- Never imply ownership or acquisition unless the source explicitly supports it.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

