Index of every page on this site: llms.txt
All skills
Run it yourself

Iterative writing loop. Gemini 3 Pro writes, Claude Agent SDK reviews autonomously. Use for blog posts, docs, technical content needing quality iteration.

  • 5 files
  • 25 KB
  • Updated 10 months ago
  • GitHub
Use this Skill: https://skilld.dev/gh/onmax/claude-config/writer

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

SKILL.md

≈41 tokens for metadata: the name and description. ≈649 when used: this file. ≈1.9k more on demand in 3 files.

Description uses 2.0% of example budget

Before choosing a Skill, your Agent reads its name and description. All available Skills share that space.

  • A shorter description leaves more room for other Skills. This entry exceeds our 1% size suggestion.

In our Claude Code example, all Skill names and descriptions share 8,000 characters. This Skill uses ≈163 characters, or 2.0%.

The 1% threshold is a size suggestion. Longer descriptions can still fit.

Your model, settings, and other Skills decide how much text your Agent can read.

Example settings and source

The example uses a 200k-token context and default Claude Code settings. The count includes the name, description, separators, and when_to_use when present. Codex also counts local file paths.

Skit's source and limits: Codex 0.160.1, Claude Code 2.1.292.

Writer Skill

Iterative writing loop that combines Gemini 3 Pro's writing capabilities with Claude's review.

Flow

  1. User provides topic/context
  2. Gemini 3 Pro writes draft
  3. Claude Agent SDK reviews autonomously
  4. If not approved, Gemini rewrites with feedback
  5. Loop until approved (max 15 iterations)
  6. DeepL improves grammar/style (if API key set)
  7. Claude merges DeepL suggestions (preserves code, markdown, technical terms)

Usage

shelve run --project Personal -- npx tsx ~/.claude/skills/writer/scripts/write-loop.ts "Write a blog post about Vue 3 composition API"

With context file:

shelve run --project Personal -- npx tsx ~/.claude/skills/writer/scripts/write-loop.ts "Write docs for this component" --context ./Component.vue

Environment Variables

Use shelve skill to manage env vars in project "Personal":

  • GOOGLE_AI_KEY_1 - Primary Google AI API key
  • GOOGLE_AI_KEY_2 - Secondary key (quota rotation)
  • GOOGLE_AI_KEY_3 - Tertiary key (quota rotation)
  • DEEPL_API_KEY - DeepL API key (for grammar step, optional)

Options

Templates

  • post - Technical blog post (default)
  • doc - Documentation
  • github-issue - GitHub issue
shelve run --project Personal -- npx tsx ~/.claude/skills/writer/scripts/write-loop.ts --template doc "Vue composable docs"

Model

  • --model pro - Gemini 3 Pro (default, faster)
  • --model pro-think - Gemini 3 Pro with thinking mode (slower, better quality)
shelve run --project Personal -- npx tsx ~/.claude/skills/writer/scripts/write-loop.ts --model pro-think "Complex technical post"

Resume

If interrupted, resume from last state:

shelve run --project Personal -- npx tsx ~/.claude/skills/writer/scripts/write-loop.ts --resume

State saved to .writer-state.json after each iteration.

Grammar Options

  • --no-grammar - Skip DeepL grammar step
  • --style, -s - Writing style: academic|business|casual|simple
  • --tone, -o - Tone: confident|diplomatic|enthusiastic|friendly
  • --lang, -l - Target language (default: en-US)
  • --max-iterations, -m - Max iterations (default: 15)

Default styles per template:

  • post: business + friendly
  • doc: simple + confident
  • github-issue: business + diplomatic
# Custom style/tone
shelve run --project Personal -- npx tsx ~/.claude/skills/writer/scripts/write-loop.ts --style casual --tone enthusiastic "Fun intro to TypeScript"

# Skip grammar step
shelve run --project Personal -- npx tsx ~/.claude/skills/writer/scripts/write-loop.ts --no-grammar "Quick draft"

Source: SKILL.md on GitHub

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

No third-party reports yet.

Signed by skilld at 0a85622. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 4 days ago.

Activeupdated 10 months ago

README badge

README badge for onmax/claude-config/writer