Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Draft a blog post from ideas and resources. Use when users want to write a blog post, create content from research, or draft articles. Guides through research, brainstorming, outlining, and iterative drafting with version control.
.claude/skills/luongnv89-blog-draft/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | 152% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 75% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 178% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 51% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 74% | 0% |
text$ARGUMENTS
You MUST consider the user input before proceeding. User should provide:
IMPORTANT: If the user is requesting updates to an existing blog post, skip steps 0-8 and start directly at Step 9. Read the existing draft file(s) first, then proceed with the iteration process.
Follow these steps sequentially. Do not skip steps or proceed without user approval where indicated.
YYYY-MM-DD-short-topic-name blog-posts/ └── YYYY-MM-DD-short-topic-name/ └── resources/
resources/ subfolder in the blog post directoryresources/ as markdown filesresources/resources/:resources/source-1-[short-name].mdresources/source-2-[short-name].mdmarkdown # Source: Title/URL]
## Key Points
## Relevant Quotes/Data
## How This Relates to Topic Brief explanation of relevance
markdown # Blog Post Outline: Title]
## Meta Information
## Proposed Structure
### Hook/Introduction
### Section 1: Title]
### Section 2: Title]
Continue for all sections...]
### Conclusion
## Sources to Cite
OUTLINE.md in the blog post folder.docs: Add outline for blog post - [topic-name]draft-v0.1.md in the blog post folder.markdown # Blog Post Title]
Optional: subtitle or tagline]
Full content with inline citations...]
---
## References
docs: Add draft v0.1 for blog post - [topic-name]If user requests changes:
draft-v[X.Y].mdIf user approves:
final.md if user requestsAll drafts are preserved with incremental versioning:
draft-v0.1.md - Initial draftdraft-v0.2.md - After first round of feedbackdraft-v0.3.md - After second round of feedbackThis allows tracking the evolution of the blog post and reverting if needed.
blog-posts/
└── YYYY-MM-DD-topic-name/
├── resources/
│ ├── source-1-name.md
│ ├── source-2-name.md
│ └── ...
├── OUTLINE.md
├── draft-v0.1.md
├── draft-v0.2.md (if iterations)
└── draft-v0.3.md (if more iterations)Last Updated: August 4, 2026 Claude Code Version: 2.1.220 Sources:
Compatible Models: Claude Fable 5, Claude Opus 5, Claude Sonnet 5, Claude Sonnet 4.6, Claude Opus 4.8, Claude Haiku 4.5
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 22,131 | 4,051 | -82% | 1 | 1 | 0% | 3,851 | 2,895 | -25% | 0 | 0 | — |
case-02 | fail→fail | 14,249 | 18,624 | +31% | 1 | 1 | 0% | 2,395 | 4,761 | +99% | 0 | 0 | — |
case-03 | fail→fail | 5,747 | 2,992 | -48% | 1 | 1 | 0% | 973 | 2,592 | +166% | 0 | 0 | — |
case-04 | fail→pass | 6,553 | 4,232 | -35% | 1 | 1 | 0% | 1,174 | 2,961 | +152% | 0 | 0 | — |
case-05 | fail→pass | 7,699 | 2,394 | -69% | 1 | 1 | 0% | 1,521 | 2,662 | +75% | 0 | 0 | — |
case-06 | pass→pass | 12,810 | 6,234 | -51% | 1 | 1 | 0% | 2,175 | 3,512 | +61% | 0 | 0 | — |
case-07 | fail→pass | 5,344 | 1,907 | -64% | 1 | 1 | 0% | 945 | 2,629 | +178% | 0 | 0 | — |
case-08 | fail→pass | 9,753 | 2,475 | -75% | 1 | 1 | 0% | 1,802 | 2,722 | +51% | 0 | 0 | — |
case-09 | fail→fail | 7,201 | 1,588 | -78% | 1 | 1 | 0% | 1,218 | 2,439 | +100% | 0 | 0 | — |
case-10 | fail→pass | 10,767 | 4,388 | -59% | 1 | 1 | 0% | 1,749 | 3,035 | +74% | 0 | 0 | — |
case-11 | fail→pass | 6,660 | 2,434 | -63% | 1 | 1 | 0% | 1,134 | 2,602 | +129% | 0 | 0 | — |
case-12 | fail→fail | 4,284 | 2,359 | -45% | 1 | 1 | 0% | 762 | 2,561 | +236% | 0 | 0 | — |
case-13 | pass→pass | 2,460 | 2,442 | -1% | 1 | 1 | 0% | 404 | 2,639 | +553% | 0 | 0 | — |
case-14 | fail→pass | 12,699 | 4,829 | -62% | 1 | 1 | 0% | 2,128 | 3,070 | +44% | 0 | 0 | — |
case-15 | fail→pass | 9,367 | 4,030 | -57% | 1 | 1 | 0% | 1,520 | 2,963 | +95% | 0 | 0 | — |
case-16 | fail→pass | 10,599 | 2,500 | -76% | 1 | 1 | 0% | 1,756 | 2,620 | +49% | 0 | 0 | — |
case-17 | pass→pass | 8,581 | 1,934 | -77% | 1 | 1 | 0% | 1,379 | 2,537 | +84% | 0 | 0 | — |
case-18 | fail→fail | 6,455 | 2,987 | -54% | 1 | 1 | 0% | 1,123 | 2,713 | +142% | 0 | 0 | — |
case-19 | fail→pass | 5,774 | 1,933 | -67% | 1 | 1 | 0% | 1,055 | 2,492 | +136% | 0 | 0 | — |
case-20 | pass→fail | 8,765 | 4,311 | -51% | 1 | 1 | 0% | 1,425 | 2,497 | +75% | 0 | 0 | — |
case-21 | pass→pass | 13,035 | 12,064 | -7% | 1 | 1 | 0% | 2,628 | 4,044 | +54% | 0 | 0 | — |
case-22 | fail→fail | 2,551 | 3,737 | +46% | 1 | 1 | 0% | 425 | 2,916 | +586% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 21 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +41 percentage points is the difference between those two pass rates over the 21 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
| Model | Method | Date | Lift |
|---|---|---|---|
| gemini-3.6-flash | verified | 8/7/2026 | +45% |
Other measured skills in the registry, with their headline benchmark lift.