Live survey304 models · 14 providers · 22,792 auditsAudited grill-with-docs · SAFE →What changedHow we verifyllms.txt

Agent skill · Audited from source

x-writer

skill · ognjengt/founder-skillsMIT licenceAudited 31 Oct 2026 · source 90bc6f1424df

★ 652pushed 165 days ago
✓SAFEGrade B · 89/100 · 31 Oct 2026 · 90bc6f1
652repo stars · whole ognjengt/founder-skills

Is it safe? Yes — SAFE, grade B (89/100). We read this skill's source at commit 90bc6f1424df and found no filesystem, network or shell access and no hard-coded secrets. What it does: Creates viral X (Twitter) posts using proven formats, post templates, and creator voice matching. Use when user needs engaging, high-performing posts for X/Twitter.

Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
More
Cite

Cite this page

ModelsAtlas. “x-writer (ognjengt/founder-skills)”. Retrieved 31 Oct 2026. https://modelsatlas.com/skills/ognjengt/x-writer

Trust audit

Audit 0.4.1 · source read 31 Oct 2026

Trust score89/100✓SAFE B
0 critical0 high0 medium0 low0 info

SAFE means nothing in the source contradicts what the skill says it does. Grade B is shared by 16,667 of 17,602 audited skills.

What the source does

  • FilesystemNone observedNo reads or writes in the source
  • NetworkNone observedNo outbound calls in the source
  • ShellNone observedNo commands run in the source
  • DependenciesPinnedNothing floats
  • SecretsNone foundNo keys in source

Five audit layers

LayerWhat we checkResultFindings
L0 ProvenanceRepo owner, licence, archive status, activityPASS0
L1 Static sourceSecrets, path traversal, supply chain, hidden codeNA0
L2 InstructionsPrompt injection, authority framing, hidden directivesPASS0
L3 InventoryThe files and declarations match the sourcePASS0
L4 SandboxRun in isolation and watch behaviourSKIPPED why0

Gates applied: no behavioural pass.

Findings (0)

No findings at any severity. We re-audit when the source changes.

Audit 0.4.1 · source sha 90bc6f1424df · 31 Oct 2026Report an issue or request a re-scan

Provenance

Who published it, and what the verdict is pinned to

  • Publisherognjengt GitHub userOwner of the repository this skill is read from
  • Source folderognjengt/founder-skills/skills/x-writerThe exact folder we audited
  • LicenceMITAs the repository states it
  • Commit pinned for this verdict90bc6f1424dfA new audit runs when the source changes
  • Repository statusNot archivedLast push 165 days ago
  • Sandboxed runNot runSkills are read, not executed; layer L4 is skipped

What it tells the agent

Full source on GitHub ↗

Its description: Creates viral X (Twitter) posts using proven formats, post templates, and creator voice matching. Use when user needs engaging, high-performing posts for X/Twitter.

---
name: x-writer
description: Creates viral X (Twitter) posts using proven formats, post templates, and creator voice matching. Use when user needs engaging, high-performing posts for X/Twitter.
---

# X Writer

## Purpose
Generate 3 viral X posts in different proven formats, matched to a creator voice, using battle-tested templates and patterns that drive engagement.

---

## Execution Logic

**Check $ARGUMENTS first to determine execution mode:**

### If $ARGUMENTS is empty or not provided:
Respond with:
"x-writer loaded, proceed with your topic or idea"

Then wait for the user to provide their requirements in the next message.

### If $ARGUMENTS contains content:
Proceed immediately to Task Execution (skip the "loaded" message).

---

## Task Execution

When user requirements are available (either from initial $ARGUMENTS or follow-up message):

### 1. MANDATORY: Read Reference Files FIRST
**BLOCKING REQUIREMENT — DO NOT SKIP THIS STEP**

Before doing ANYTHING else, you MUST use the Read tool to read ALL three reference files. This is non-negotiable:

```
Read: ./references/formats.md
Read: ./references/posts.md
Read: ./references/voices.md
```

**What you will find:**
- **formats.md**: 8 proven post formats with structure templates, psychology, and format-specific rules
- **posts.md**: 49+ proven viral posts organized by format type — the example library
- **voices.md**: Creator voice profiles with writing DNA, signature patterns, and example posts

**DO NOT PROCEED** to Step 2 until you have read all three files and have their content in context.

### 2. Check for Business Context
Check if `FOUNDER_CONTEXT.md` exists in the project root.
- **If it exists:** Read it and use the business context to personalize output (industry terminology, audience pain points, brand voice, company name, products).
- **If it doesn't exist:** Proceed using defaults from "Defaults & Assumptions."

### 3. Analyze Input & Select Voice
From the user's requirements, extract:
- **Topic/idea** — What they want to post about
- **Goal** — What they want the post to achieve (engagement, authority, education, humor)
- **Voice preference** — If they specified a voice or tone

**Voice Matching Logic:**

If the user specified a voice → use that voice directly.

If the user did NOT specify a voice:
1. Analyze the topic and FOUNDER_CONTEXT.md brand voice
2. Auto-suggest the best match based on topic:
   - Business advice, mindset, pricing, sales, productivity → suggest **Alex Hormozi**
   - Life philosophy, contrarian takes, self-improvement, minimalism → suggest **Naval Ravikant**
   - Startup advice, CEO/founder roles, hiring, company building → suggest **Andrew Gazdecki**
   - Self-improvement, productivity hacks, fitness + learning, habits → suggest **Dakota Robertson**
   - AI trends, tool launches, tech stack recommendations, urgent opportunities → suggest **Machina**
   - SaaS growth, AI productivity, founder mindset, revenue strategy → suggest **Ognjen Gatalo**
   - Personal responsibility, escaping default paths, harsh reality checks, focus training → suggest **Dan Koe**
3. Ask the user to confirm using AskUserQuestion:
   - **[Suggested Voice] (Recommended)** — [one-line description of why it fits]
   - **[Other Voice 1]** — [one-line description]
   - **[Other Voice 2]** — [one-line description]
   - **My own voice** — Natural tone based on FOUNDER_CONTEXT.md brand voice

### 4. Generate 3 Viral Posts
Using the formats, posts, and voice you loaded in Steps 1-3:

1. **Select 3 DIFFERENT formats** from formats.md that best fit the user's topic and goal
2. **Study the example posts** in posts.md for those formats — internalize the rhythm, structure, and length
3. **Apply the selected voice** from voices.md — match their sentence structure, rhythm, and signature patterns (not their exact words or examples)
4. **Draft each post** following the format's structure template, the voice's DNA, and all Writing Rules below
5. **Run each post through the Screenshot Test:** "Would someone screenshot this and share it?" If no → rewrite before continuing

**Critical requirements:**
- Each post must use a DIFFERENT format (no repeats)
- Each post must sound like the selected voice (structural match, not word-for-word copy)
- Each post must be about the user's SPECIFIC topic (not generic advice)
- Each post must follow the format's structure template from formats.md
- Each post must be ready to copy-paste and post immediately — no placeholders

### 5. Format and Verify
- Structure output according to **Output Format** section
- Complete **Quality Checklist** self-verification before presenting output
- If any post feels generic, forced, or wouldn't pass the Screenshot Test → rewrite before presenting

---

## Writing Rules
Hard constraints. No interpretation.

### Core Rules (Apply to ALL posts)
- Maximum 280 characters for single posts. If a post exceeds 280 characters, it must be structured as a thread (each tweet under 280 characters).
- First line is the hook — it determines whether people click "Show more." Make it the strongest line in the post.
- Use line breaks strategically. White space = visual breathing room = readability.
- No hashtags. They reduce reach and look amateur on X.
- No engagement bait ("like if you agree", "RT if this resonates", "thoughts?"). It kills reach.
- No emojis unless the voice or topic specifically calls for it. When used, maximum 1-2 per post.
- Specific numbers > vague claims ("17 strategies" not "many strategies").
- Opinions > facts. X rewards bold takes, not Wikipedia summaries.
- Active voice only. Never passive.
- Present tense preferred.
- Every word must earn its place. If you can cut a word without losing meaning, cut it.
- No "I think" or "In my opinion" — just state the opinion as fact. Confidence is voice.

### Voice-Matching Rules
- Match the voice's SENTENCE STRUCTURE — short choppy lines vs flowing sentences, fragments vs complete thoughts.
- Match their RHYTHM — how they 

Excerpt, first 6,000 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.

Install

Pinned to the audited commit 90bc6f1

Download folder ↗
git clone https://github.com/ognjengt/founder-skills
git -C founder-skills checkout 90bc6f1424df

The author documents no install command, so this uses the repository directly. Our verdict covers commit 90bc6f1424df only.

Works in which agents?

What its own text says, not a test

AgentStatusMore
Claude CodeMentioned in its text, not testedSkills for Claude Code · About Claude Code
CodexMentioned in its text, not testedSkills for Codex · About Codex
CursorMentioned in its text, not testedSkills for Cursor · About Cursor
“Mentioned” means the skill’s text names the agent. We have not run it there.

Audit history

Every audit, every source change

  1. ✓SAFE B 89 / 100

    first audit · source 90bc6f1424df

README badge

For maintainers · shows the latest audit

ModelsAtlas: SAFE BMODELSATLASSAFE · B
[![ModelsAtlas: SAFE B](https://modelsatlas.com/skills/ognjengt/x-writer/badge.svg)](https://modelsatlas.com/skills/ognjengt/x-writer/)

The badge shows our verdict and grade for this skill, and links back to this page. It changes when the audit does.

Add this badge to your README

[![ModelsAtlas: SAFE B](https://modelsatlas.com/skills/ognjengt/x-writer/badge.svg)](https://modelsatlas.com/skills/ognjengt/x-writer/)

Questions about x-writer

What does the x-writer skill do?
Claude skills for founders
Is x-writer safe to install?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean skill reads B.
What can x-writer access on my machine?
The audit observed no filesystem, network or shell use at all in its source.
Which assistants does x-writer work with?
Its documentation mentions claude-code, codex and cursor. That is what the text claims, not a compatibility test we ran.
How current is this page?
The grade is for one exact copy of the source (90bc6f1424df), read on 2026-10-31. The repository is watched, and a new audit runs when it changes — this is the first audit.

Sources

Spotted something wrong? Report a correction · Methodology

Keep going in your AI

Opens with this page as the source. We add nothing else to the prompt.