Skip to content
SkillModel-led2,786 tokens in use

Humanize Korean for Codex

The Codex CLI edition of Humanize Korean: it diagnoses AI-written Korean, rewrites it on a light, standard or heavy path and gates the result with scripts.

6,692 stars on the repository MIT Licenseepoko77-ai/im-not-ai · codex/skills/humanize-korean
git clone https://github.com/epoko77-ai/im-not-ai.git && cd im-not-ai ./install.sh --codex-only
View source

Install command as written in the repository's README

Skill label

measured from its files

What installing Humanize Korean for Codex costs

254tokens idle
2,786tokens in use

Idle is what sits in context while the skill is installed. In use is its instructions, loaded when it runs. Counts are approximate, made offline with cl100k_base.

Model dependence Model-led

Script-ledMixedModel-led

Mostly instructions, so how well it works depends on the model that follows them.

This describes the skill's structure, how much of it is scripts. It does not say how well it works. Scripts: scripts/prepare_monolith_input.py, scripts/reassemble_chunks.py, scripts/verify_change_rate.py, scripts/verify_gates.py.

Touches

shelldetectedwrites filesdetected

Declared tags come from the skill's frontmatter; detected ones from its scripts, its code blocks and the commands it names in inline code. This is detected, not audited.

How it scores

Quality

How well the skill is written, counted from its own files: its frontmatter, its steps, examples, stated output and constraints, its size and whether every file it ships is one it names.How we score it →
Fair
  • Frontmatter8/15
  • Structure16/25
  • Examples12/15
  • Output0/15
  • Size8/15
  • Resources3/15

Security

Starts at 100 and loses points for each thing in the skill's files that a reader should look at first.How we score it →
Good
  • Runs shell commands−5

    does this on the machine, and the frontmatter does not say so

    SKILL.md, line 24
  • Writes files−8

    does this on the machine, and the frontmatter does not say so

    SKILL.md, line 24
  • File that was not read−15

    a file the skill ships that was not read, so what it does is unknown

    references

Repository health

Pre-computed score based on development activity, maintenance, community, maturity, and trend momentum.How we score it →
Excellent
  • Development activity96
  • Maintenance100
  • Community64
  • Maturity28
  • Momentum40

This is the Codex edition of Humanize Korean. It rewrites Korean text produced by AI so that translated-sounding phrasing, stock phrases, mechanical structure and uniform rhythm are removed while facts, numbers, names and quotes stay as written.

It uses Codex collaboration agents for diagnosis, rewriting and final review when they are available, and runs the same roles one after another in the main agent when they are not. Role contracts live in reference files, and Python scripts score the input and check the output.

What You Get

  • A rewritten Korean text saved as final.md with one summary block at the end
  • A run folder holding the input, metrics, diagnosis and finalize result
  • A status line with route, change rate, grade and self-check results
  • Follow-up modes for a second pass, one category, one paragraph or a different strength

Common Use Cases

  • Rewriting an AI-written Korean article from inside Codex CLI
  • Forcing the full diagnose, rewrite and finalize path with a strict mode request
  • Re-running only one paragraph or one pattern category after a first pass

What Makes It Unique

This edition is written for Codex. It resolves its own folder through a physical path, reads role contracts from reference files and works with or without Codex collaboration agents. It keeps the full light, standard and heavy routing, which the Copilot and Gemini editions of the same project leave out.

Under The Hood

The skill runs shell commands: a Python script that scores the input and writes a metrics file, a script that joins chunks, and a gate script that checks the change rate and structure. It creates and writes files in a workspace folder under the current directory, and it can start Codex collaboration agents for the diagnosis, rewrite and finalize roles. The script files in this skill folder are links to scripts kept elsewhere in the repository. It does not fetch web pages or call outside APIs.

Join founders building with open source

Opinionated takes, migration guides, cost-saving tips, and insights from the open source ecosystem.

Subscribe on SubstackJoin 750+ subscribers