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 / /content command dispatches the content-writer agent to generate three platform-native draft variations conditioned on the active persona. Provide a target platform as the argument and the agent will produce Draft 1, Draft 2, and Draft 3, each adapted to the platform’s native patterns, CTA conventions, and the persona’s voice. All drafts are saved to disk before the response is returned.

​
Invocation

$content x
$content linkedin
$content naver-blog
The $ARGUMENTS value is the target platform. Accepted values:
ValuePlatform
xX (Twitter)
threadsThreads
linkedinLinkedIn
naver-blogNaver Blog
brunchBrunch

​
What it does

  • Reads personas/_active.md to resolve the active persona directory.
  • Loads personas/{active}/persona.md, nuance.md, and accounts.md to build the full persona context.
  • Passes the platform argument and current topic context to the content-writer agent.
  • Returns 3 platform-native draft variations labelled Draft 1, Draft 2, and Draft 3.
  • Saves all generated drafts to content/drafts/ before returning the response.
  • Re-reads the saved draft file and includes the confirmed saved path in the response.
  • If no file was actually written, the content-writer reports failure rather than claiming draft persistence.
The three drafts are variations, not rankings. Each explores a different angle, opening hook, or CTA style within the same platform and topic. You choose which one to take to QA.

​
Prerequisites

All three persona files must be present before generating content: 
personas/{active}/persona.md    ← core identity (from $interview)
personas/{active}/nuance.md     ← voice and tone (from structuring engine)
personas/{active}/accounts.md   ← platform account context (from structuring engine)
Run $interview to build a persona, then $trend to identify a topic, before calling $content [platform]. While a trend scan is not strictly required, content quality is significantly higher when it has a topic signal to work with.

​
Output

Drafts are saved to: 
content/drafts/
The response includes the confirmed file path and all three drafts inline. Example structure: 
Saved: content/drafts/2025-07-01_x_draft.md

---
Draft 1
[platform-native content variation 1]

Draft 2
[platform-native content variation 2]

Draft 3
[platform-native content variation 3]

​
Guardrails & Notes

The content-writer must write a file before it reports success. If you receive a response without a confirmed content/drafts/ path, the save failed — re-run the command.
  • Platform argument is required. Calling $content without a platform will result in an error or undefined behaviour.
  • The content-writer reads accounts.md for platform account context. If this file is missing, drafts may lack account-specific nuance (handle, bio alignment, follower signal).
  • Draft files are used as the default input by $qa. The most recently saved file in content/drafts/ is picked up automatically.
The $content skill is executed by the content-writer agent. See the Content-writer agent page for details on persona-conditioned drafting, platform pattern adaptation, and CTA shaping.

Build docs developers (and LLMs) love

Get started for freeTalk to us