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.

Polysona ships a Claude Code plugin manifest in .claude-plugin/. Rather than relying on file auto-discovery, you register a local marketplace entry and install the plugin explicitly — after that, all eight persona commands are available as /slash commands in any Claude Code session you start in the repo directory.

​
Plugin Manifest

The plugin is defined in .claude-plugin/plugin.json. It declares the plugin name, version, the eight skill paths, and the five agent definition files:
.claude-plugin/plugin.json
{
  "name": "polysona",
  "version": "1.3.0",
  "author": {
    "name": "LilMGenius"
  },
  "description": "Build and run multiple personas across any AI agent.",
  "skills": [
    "./skills/interview",
    "./skills/introduce",
    "./skills/trend",
    "./skills/content",
    "./skills/qa",
    "./skills/publish",
    "./skills/status",
    "./skills/export"
  ],
  "agents": [
    "./agents/profiler.md",
    "./agents/trendsetter.md",
    "./agents/content-writer.md",
    "./agents/virtual-follower.md",
    "./agents/admin.md"
  ]
}
The local marketplace manifest at .claude-plugin/marketplace.json points Claude Code’s plugin installer to the local source:
.claude-plugin/marketplace.json
{
  "name": "polysona-marketplace",
  "owner": {
    "name": "LilMGenius",
    "url": "https://github.com/LilMGenius/polysona"
  },
  "plugins": [
    {
      "name": "polysona",
      "source": "./",
      "category": "productivity"
    }
  ]
}

​
Setup

1

Clone the repository

git clone https://github.com/LilMGenius/polysona.git
cd polysona
2

Install dependencies

bun install
3

Register the local marketplace

This tells Claude Code where to find the Polysona plugin source.
claude plugin marketplace add ./.claude-plugin/marketplace.json
4

Install the plugin

claude plugin install polysona
5

Start a Claude Code session

Open Claude Code in the repo directory. The plugin skills and agents are now loaded.
6

Begin the persona pipeline

Run /interview to start the 10-framework psychology interview.
/interview

​
Command Reference

CommandAgentPurpose
/interviewprofilerRun the 10-framework psychology interview to extract persona data
/introduce—Inject the active persona into the current session context
/trendtrendsetterScan for domain trend topics relevant to your persona
/content [platform]content-writerGenerate platform-specific drafts conditioned on your persona
/qavirtual-followerSimulate virtual followers and surface the TOP 5 draft recommendations
/publishadminPublish selected content and record engagement metadata
/status—Display the active persona and current pipeline state
/export—Export persona-derived instructions into generated files for another workspace
The typical content cycle is: /interview → /introduce → /trend → /content [platform] → /qa → /publish.

​
Lifecycle Hooks

Polysona declares three lifecycle hooks in hooks/hooks.json. Whether these are executed automatically depends on your Claude Code host version’s support for hook manifests.
hooks/hooks.json
{
  "description": "Polysona lifecycle hooks for persona management",
  "hooks": {
    "SessionStart": [
      {
        "hooks": [
          { "type": "bash", "command": "bash hooks/session-start.sh" }
        ]
      }
    ],
    "PreToolUse": [
      {
        "hooks": [
          { "type": "bash", "command": "bash hooks/pre-tool-use.sh" }
        ]
      }
    ],
    "PostToolUse": [
      {
        "hooks": [
          { "type": "bash", "command": "bash hooks/post-tool-use.sh" }
        ]
      }
    ]
  }
}

SessionStart

hooks/session-start.sh — Reads personas/_active.md to find the active persona, prints the first 10 lines of persona.md, and echoes the four core PLOON rules. If no persona exists, it prompts you to run /interview.

PreToolUse

hooks/pre-tool-use.sh — When the Write tool targets personas/*, prints a warning to read the file first and to append to the interview-log section only — never overwrite the compressed PLOON core.

PostToolUse

hooks/post-tool-use.sh — Scans tool output for AI verbosity patterns (certainly, absolutely, as an AI, etc.) and warns when they are detected.
Hook auto-execution depends on Claude Code host support. Do not assume hooks run automatically in every environment. You can run them manually (bash hooks/session-start.sh) to test.

​
Codex vs Claude Code

The two integrations use different installation mechanisms and command prefixes, but expose identical pipeline capabilities.
How it loads: AGENTS.md is auto-recognized by Codex in the repo root. Skills are discovered automatically from .agents/skills/ (kept in sync via scripts/sync-codex-skills.mjs).Command prefix: $
$interview
$introduce
$trend
$content [platform]
$qa
$publish
$status
$export
Setup steps:
  1. Clone the repo and run bun install
  2. Run bun run codex:skills:sync (or node ./scripts/sync-codex-skills.mjs) to mirror skills/ into .agents/skills/
  3. Open Codex in the repo directory — AGENTS.md and skills are auto-loaded
Re-sync required: Yes, after editing any file under skills/. Restart Codex if an updated skill does not appear.

Build docs developers (and LLMs) love

Get started for freeTalk to us