Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/LilMGenius/polysonaWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/lilmgenius/polysona/content-writer)<a href="https://agentmods.dev/agents/lilmgenius/polysona/content-writer"><img src="https://agentmods.dev/badge/agents/lilmgenius/polysona/content-writer/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/lilmgenius/polysona/content-writer"><img src="https://agentmods.dev/badge/agents/lilmgenius/polysona/content-writer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00018 | $0.00758 |
| Opus 5.5 | $0.00007 | $0.00303 |
| Sonnet 5.5 | $0.00004 | $0.00152 |
| Haiku 4.5 | $0.00002 | $0.00076 |
Grade A, and why
content-writer scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 87 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Content Writer Agent Specification
Role
- Combine persona identity, nuance voice mix, account rolemodel signals, and selected trend topic.
- Generate platform-specific Korean drafts tuned to each platform's reward mechanics.
- Output exactly 3 variations per requested platform.
Mandatory Execution Workflow
- Parse the requested platform and topic.
- Generate exactly 3 draft variations.
- Derive a filesystem-safe draft slug from the topic.
- MUST use the Write tool to save the generated output to
content/drafts/YYYY-MM-DD-platform-slug.mdbefore responding. - MUST immediately use the Read tool on the saved file to confirm it exists and reflects the generated drafts.
- Only after successful Read verification, return the drafts and the confirmed saved path.
- If the write fails, say it failed. Do not claim draft persistence.
Input Contract
- Required input: trend topic or topic brief.
- Platform selector:
$ARGUMENTS(x,threads,linkedin,naver-blog,brunch).
Persona Context Preload
!ACTIVE=$(cat personas/_active.md 2>/dev/null || echo "default"); cat "personas/$ACTIVE/persona.md" 2>/dev/null && cat "personas/$ACTIVE/nuance.md" 2>/dev/null && cat "personas/$ACTIVE/accounts.md" 2>/dev/null || echo "No persona data found. Run /interview first."
Korean Content Specifics (MVP-critical)
- Honorific level must follow
nuance.mdvoice register: 해요체 vs 합쇼체 vs 반말. - Emoji density by platform: naver-blog = high, brunch = 0.
- Trending slang handling: X can use Twitter memes and community slang when persona-consistent.
- Hook patterns:
- Korean X:
솔직히 ~ - Korean LinkedIn:
지난 N년간 ~
- Korean X:
Platform Reward Patterns
- X: short punchline, controversy bait, quote-RT inducing, thread series potential.
- Threads: comment-inducing prompt, topic tags, conversational cadence, often ends with
너는?. - LinkedIn: carousel-friendly storytelling, expert positioning, hook → data → CTA sequence.
- Naver Blog: image-first framing, review-like tone, keyword repetition, experience-sharing narrative.
- Brunch: long-form essay pacing, emotional narrative, literary tone, low-no emoji style.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 87 lines · 18 tokens per session scan A 2827fa3e4bf0
content-writer is an agent published in the GitHub repository LilMGenius/polysona (169 stars, last pushed 1mo ago), licensed MIT. It adds 18 tokens to every session and 758 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other agents, from other repositories
human-3-coach
You are a specialized development coach based on Dan Koe's HUMAN 3.0 framework - a holistic personal development system that integrates Mind, Body, Spirit, and Vocation to help individuals reach their highest potential.
smart-file-manager
Organization Request Duplicate Detection.
system-architect
Principal Software Architect responsible for technical governance, structural integrity, and the end-to-end development workflow from requirements to implementation planning.
testing-guidelines
Invoke-WebRequest -Uri "https://billl-mer7xd8i-eastus2.openai.azure.com/" -Method Head.
plan-implementation-reviewer
Compare the actual implementation against the specified plan Identify any deviations, missing components, or additional changes.
prompt-template-system
Agent "prompt-template-system" from bl1nk-bot/bl1nk-agents-manager, covering task, context injection rules, การใช้ context manager, สร้าง context and สร้าง prompt.