ghost-user

A first-time-user walkthrough that examines a product or flow from the viewpoint of someone with no background knowledge.

In plain words
What is it for?
Use it to test onboarding and first contact with an app or workflow, describing what a new user notices, assumes, and cannot understand.
Why use it?
It reveals confusing instructions, invisible steps, and assumptions that experienced builders may overlook.

Skill for Claude CodeCodex

Install

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.

agentmods
npx agentmods add skills/pyros-projects/limitless/ghost-user
Any agent
npx skills add pyros-projects/limitless --skill ghost-user
Clone the repo
git clone --depth 1 https://github.com/pyros-projects/limitless

Made for: Claude Code, Codex.

Per session 101 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 840 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00101 $0.00840
Opus 5 $0.00051 $0.00420
Sonnet 5 $0.00020 $0.00168
Haiku 4.5 $0.00010 $0.00084

Measured yesterday against content hash 0fe5a1d69de1, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

ghost-user 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 yesterday.

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.

plugins/after-hours/skills/ghost-user/SKILL.md · 105 lines

How it starts

The opening of the file, as written. The whole thing — 105 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Ghost User

Overview

This skill simulates first contact.

Not a power user. Not a chaos monkey. Not a formal usability study.

Just a quiet first-time user moving through the surface with no internal context and no special sympathy for what the builders meant.

When to Use

Use this when:

  • a flow feels "obvious" and you want to verify that
  • onboarding is thin or absent
  • the product may be assuming too much prior knowledge
  • you want a first-contact honesty pass without setting up full user research

Do not use this to invent edge cases for sport. Stay close to plausible first-use reality.

If you have already read project docs, architecture notes, or implementation details in the current session, explicitly say that your ghost-user pass is partially contaminated and flag which observations may be insider-influenced.

What to Produce

Walk through the product or flow as the ghost user would:

  • what they notice first
  • what they assume
  • what confuses them
  • what feels invisible
  • what requires author knowledge to decode

Narrate it plainly. The power of this skill is not sophistication. It is honesty.

Working Loop

  1. Forget the author view. Start from the surface alone. Do not read README files, docs, architecture notes, or source code before the pass if the raw product surface is available.

  2. Walk the first contact honestly. What do I see? What do I think this means? What do I try next?

  3. Notice the invisible assumptions. Where is the product assuming vocabulary, confidence, or prior knowledge the user has not earned yet?

  4. End with the highest-leverage fixes. Do not drown the user in notes. Point to the few changes that would most improve first contact.

Suggested Output Shape

  • What I notice first
  • What I assume
  • Where I get confused
  • What required inside knowledge
  • The 3 highest-leverage fixes

Tiny Example

  • What I notice first: I see a button called Init.
  • What I assume: Maybe it creates something, but I do not know what.
  • Where I get confused: Clicking it changes the page, but nothing tells me what happened.
  • What required inside knowledge: Knowing that Init means "create a new workspace."
  • The 3 highest-leverage fixes: Rename the button, add immediate confirmation, and show the created thing in context.

Read the full file on GitHub · 105 lines

Files

What ships with it

1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

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.

  1. yesterday First seen · 105 lines · 101 tokens per session scan A 0fe5a1d69de1

Subscribe to this mod's changes

ghost-user is a skill published in the GitHub repository pyros-projects/limitless (9 stars, last pushed 19d ago), licensed MIT. It adds 101 tokens to every session and 840 once invoked, about $0.0005 per session on Opus 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 2026-08-31.