Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/LilMGenius/polysona/llms.txt

Use this file to discover all available pages before exploring further.

The Content-Writer is Polysona’s drafting engine. It takes a trend topic and a platform target, then combines your persona’s motivational signature, voice register from nuance.md, and rolemodel style cues from accounts.md to produce three distinct draft variations. Each draft is written for a specific platform’s native reward mechanics — not as a generic adaptation, but as a first-principles rewrite conditioned on how that platform distributes attention. The three-variation structure ensures you always have options before handing off to the virtual-follower for QA.

​
Configuration

FieldValue
namecontent-writer
toolsRead, Write
Codex command$content [platform]
Claude Code command/content [platform]

​
Role

The Content-Writer performs three jobs:
  1. Persona-conditioned drafting — anchors tone, motivation, and decision style to persona.md (core, decide, energy, blind sections).
  2. Platform pattern adaptation — rewrites the same topic intent for each platform’s native hook and cadence patterns.
  3. CTA shaping — closes each draft with a call-to-action appropriate to the platform and persona voice.

​
Input Context Loaded

Before generating, the Content-Writer loads three files from the active persona:
FileSections Used
persona.mdcore, decide, energy, blind
nuance.mdvoice (honorific level, phrasing constraints, avoid-list), platform
Trend inputTopic from the latest content/trends/ scan

​
Supported Platforms

Pattern: Short, sharp, controversy-bait, quote-RT inducing, thread series potential.Emoji density: Low.Korean hook pattern: 솔직히 ~ (Honestly, ~)Reward mechanic: First line must stop the scroll. Punchline structure or sharp provocation. Content should have standalone value but invite a reply or repost.

​
Draft Generation Rules

The Content-Writer produces exactly 3 draft variations per invocation. All three share the same topic intent but are rewritten for platform-native behavior. Each draft includes:
  • Hook — the opening line designed to stop scrolling or draw the reader in
  • Body — the main content adapted to platform cadence
  • CTA — a platform-appropriate call to action
  • Platform-fit note — a short rationale for the draft’s structural choices
Across all three drafts, the Content-Writer:
  • Reflects persona motivation and decision style from persona.md
  • Respects phrasing constraints and the avoid-list from nuance.md
  • Uses honorific level from nuance.md voice register (해요체, 합쇼체, or 반말)
  • References rolemodel style patterns from accounts.md without copying wording

​
Mandatory Execution Workflow

1

Parse platform and topic

Read the platform argument from $ARGUMENTS and confirm the topic from the latest trend scan or the user-provided brief.
2

Load persona context

Read persona.md, nuance.md, and accounts.md from the active persona directory.
3

Generate 3 draft variations

Produce exactly 3 drafts for the specified platform, each with hook, body, CTA, and platform-fit note.
4

Derive a filesystem-safe draft slug

Generate a slug from the topic to use as the output filename.
5

Write draft file (required)

Use the Write tool to save the generated output to:
content/drafts/YYYY-MM-DD-platform-slug.md
This step is mandatory. Do not proceed without writing the file.
6

Verify the written file

Immediately use the Read tool on the saved file to confirm it exists and reflects the generated drafts. Only after successful verification, return the drafts and the confirmed saved path.
7

Report on failure

If the write fails, report it explicitly. Do not claim draft persistence.

​
Draft File Template

# Drafts: <PLATFORM> — <SLUG>

## metadata

platform     | <platform>
generated_at | <YYYY-MM-DD>
persona      | <persona-id>
topic        | <topic>

## draft_1

<draft 1 — hook, body, CTA, platform-fit note>

## draft_2

<draft 2 — hook, body, CTA, platform-fit note>

## draft_3

<draft 3 — hook, body, CTA, platform-fit note>

​
Example Invocations

$content x

/content x
After the Content-Writer confirms the saved draft path, hand off to the virtual-follower with $qa / /qa. The virtual-follower reads from content/drafts/ automatically.

Build docs developers (and LLMs) love

Get started for freeTalk to us