Skip to content
SkillScript-led2,103 tokens in use

Content Ops

Auto-assemble a panel of 7 to 10 experts for any copy, strategy or chart, score it, revise for up to three rounds toward 90, and learn from your rejections.

3,862 stars on the repository MIT Licenseericosiu/ai-marketing-skills · content-ops
npx skills add ericosiu/ai-marketing-skills -s content-ops
View source

Install command as written in the repository's README

Skill label

measured from its files

What installing Content Ops costs

28tokens idle
2,103tokens in use
6,201tokens, up to

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

Model dependence Script-led

Script-ledMixedModel-led

Most of this skill is scripts, so it behaves much the same whichever model runs it.

This describes the skill's structure, how much of it is scripts. It does not say how well it works. Scripts: scripts/content-quality-gate.py, scripts/content-quality-scorer.py, scripts/content-transform.py, scripts/editorial-brain.py, scripts/quote-mining-engine.py.

Touches

shelldetectednetworkdetectedwrites filesdetectedcredentialsdetected

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 →
Excellent
  • Frontmatter12/15
  • Structure25/25
  • Examples15/15
  • Output10/15
  • Size15/15
  • Resources10/15

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 →
Fair
  • File that was not read−15

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

    .env.example
  • Runs shell commands−5

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

    SKILL.md, line 11
  • Writes files−8

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

    scripts/content-quality-gate.py, line 130
  • Reads credentials−15

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

    scripts/content-transform.py, line 346
  • Reaches the network−10

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

    scripts/content-transform.py, line 595

Repository health

Pre-computed score based on development activity, maintenance, community, maturity, and trend momentum.How we score it →
Good
  • Development activity88
  • Maintenance44
  • Community56
  • Maturity28
  • Momentum40

Content Ops is a general-purpose scoring and improvement engine for work you are about to publish. It reads what you want scored, builds a panel of seven to ten experts matched to the content type and offer, picks a scoring rubric, and loops: score, list the top three weaknesses, revise, score again. The target is 90 out of 100 within three rounds.

Two experts are always on the panel: an AI writing detector weighted 1.5 times, and a brand voice match. When you reject something the panel passed, the skill records a learned pattern with a point dock, so known-bad habits are penalized in future rounds. The folder also ships scripts for quality gating, scoring, repurposing, clip discovery and quote mining.

What You Get

  • A winner with its score, the number of rounds and the panel that judged it
  • Every scoring round with expert scores, top weaknesses and the changes made
  • Runner-up scores when you compare variants
  • A source improvement brief when the content came from another skill
  • A growing patterns file of rejection and preference rules with point docks
  • Python scripts for a quality gate, a scorer, content transformation, clip discovery and quote mining

Common Use Cases

  • Raise a LinkedIn post, email or newsletter to a high score before publishing
  • Pick the best of three landing page or ad variants and iterate on the leader
  • Score a strategy memo or recommendation against a strategic quality rubric
  • Evaluate candidates with an evaluation rubric
  • Feed another skill’s output through the panel and send improvement notes back to it

What Makes It Unique

The panel is built per task from pre-built expert files and domain experts you add, and the humanizer counts for 1.5 times in the aggregate. It shows every round so the iteration trail is part of the result, and it remembers your overrides as dockable patterns. If three rounds fall short it returns the best version with its real score and what is holding it back.

Under The Hood

The method is prompt-led with reference files for experts and scoring rubrics, plus Python scripts in the folder. On start the skill runs two telemetry Python scripts that do a version check and a one-time opt-in, with local usage logs under a home folder and remote telemetry only if you opt in; those scripts are not in this skill folder. The bundled scripts call the Anthropic API with your key, download YouTube subtitles with yt-dlp, read RSS feeds and your notes folders, and write JSON files into a data folder. The skill itself also edits its own patterns file when you reject output.

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