Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    human content agent

    1

    You get finished copy plus a paired reference file under your project (default: `content/`).

    Free

    2 uses3.0 (1 review)
    human content agent

    human content agent

    Example session with this skill installed

    Write a 300-word blog post about why developers should use SQLite for small side projects. Save it to content/sqlite-post.md. Tone is warm and conversational.

    • Read your context and instructions
    • Compiled the human content agent

    Request
    /human-content-agent please create 10 jingles pick any random title

    Result
    Assumptions — playful tone; short radio/social jingles; fictional brand titles.

    1. Copper Kettle Coffee — Wake up slow, or wake up right— / Copper Kettle, dark and bright. / One small cup, then clear your head. / Copper Kettle. Go get fed.
    2. Pocket Map Shoes — Left foot, right foot, city street— / Pocket Map Shoes know your beat…
      (…full set of 10 in DEMO-paste.txt…)
    3. Quiet Hour Headphones — City noise can wait outside. / Quiet Hour—soft, closed tight. / Press play once. The world thins out. / Quiet Hour. That’s what we’re about.

    Saved: content/ten-jingles.md + content/ten-jingles-reference.md

    Screenshot: Quillbot Detect AI on this sample — 0% likely AI / 100% human-written (one scan; detectors vary).

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Human Content Agent

    Turns a plain content request into finished writing that reads like a thoughtful human wrote it—specific, natural, and usable.

    What it does

    • Writes blog posts, articles, social posts, emails, newsletters, product copy, ads, jingles, and mixed multi-format packs
    • Chooses among formal, warm, playful, or direct tone (defaults to warm and conversational when you don’t say)
    • Strips common AI tells (empty hype, stock phrases, triple-list padding, vague generalities)
    • On edits: keeps your meaning and voice; improves texture instead of rewriting everything
    • Saves a companion *-reference.md next to each content file (assumptions, claims to verify, sources)
    • Includes AdSense-oriented guidance for original, useful, policy-aware site content

    How it works

    1. You ask for content (topic + format, or attach a draft).
    2. The agent applies voice rules and format playbooks from bundled references.
    3. You get finished copy plus a paired reference file under your project (default: content/).

    Where it stops

    • Does not invent facts, stats, quotes, or sources
    • Does not impersonate real people unless you state that purpose
    • Does not guarantee AdSense approval (Google decides; this skill helps prep original content and flags risks)
    • Does not replace your brand guidelines—bring those if you have them

    Best for

    Creators, marketers, and publishers who use AI for drafting but want output that still sounds human—and who want a clear paper trail of assumptions and claims to check.

    Requires file write access. Will not fabricate facts or quotes. Direct impersonation of real people is restricted unless specified.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    3.0
    1 review
    5
    0
    4
    0
    3
    1
    2
    0
    1
    0

    This skill has been used 2 times.

    Trust & safety

    Security scan

    Security scan passed 3 days ago

    • Free with an Agensi account

    Permissions

    Read Files
    Write Files
    Listed3 days ago

    What's inside

    Frequently Asked Questions

    This skill utilizes a proprietary "negative constraint" framework that proactively filters out common AI clichéd structures, predictable transitions, and robotic buzzwords that typically trigger AI detection filters.

    The 'Human Content Agent' is a file-based skill compatible with Cursor, Claude Code, and Codex CLI, functioning directly within your local development environment or workspace.

    For every piece of content generated, the skill automatically creates a separate metadata file that flags specific claims, assumptions, and statistics that require your manual verification.

    Yes, while the skill is optimized for original creation, you can provide an existing draft and instruct the agent to refine the rhythm and clarity while maintaining your specific brand voice.

    The skill requires file system write access to save your content and the accompanying reference logs; it also requires a compatible LLM environment like Claude or GPT-4 via your chosen agent.

    New in Writing & Content