# Prompt

> Use when writing, generating, refining, reviewing, or being asked for ANY prompt aimed at an LLM (Claude / GPT / Gemini / Codex), or when building GPTs / Gems system instructions, or when a task is fact-check / research / image-generation prompting. Produces structured, framework-grounded prompts instead of ad-hoc ones. Invoke this BEFORE handing any hand-written prompt to a model.

- **Type:** Skill
- **Install:** `agentstack add skill-treylom-thiscodex-prompt`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [treylom](https://agentstack.voostack.com/s/treylom)
- **Installs:** 0
- **Category:** [AI & ML](https://agentstack.voostack.com/c/ai-and-ml)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [treylom](https://github.com/treylom)
- **Source:** https://github.com/treylom/ThisCodex/tree/master/skills/prompt

## Install

```sh
agentstack add skill-treylom-thiscodex-prompt
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# prompt — structured prompt generation

Vendored, unmodified-in-substance copy of the public **`prompt-engineering-skills`**
package (MIT). It turns a rough intent into a structured, framework-grounded
prompt rather than an ad-hoc one, and routes by task type.

## When to use (mandatory-invocation)

Invoke this skill — do not free-hand a prompt — whenever the work is:

- "write / make / improve / review a prompt" for any model
- building **GPTs** or **Gems** system instructions
- fact-check / research prompting (IFCN-grounded base templates required)
- image-generation prompting (5-stage framework)
- any request whose deliverable is itself a prompt

A bot whose config mandates `/prompt` (see the repo's
`docs/SETUP-CONFIG-GUIDE.md` §"force-invoke") MUST route here first.

## How it routes (progressive disclosure)

This `SKILL.md` is the router. Load the heavy reference only for the matched
task — do not read all of it up front:

See `references/` for detailed procedures only when the matched prompt task asks for them.

| Task | Load |
|---|---|
| Generate / refine a prompt (main workflow) | [`references/prompt-generator.md`](references/prompt-generator.md) |
| Framework / technique deep reference | [`references/prompt-engineering-guide.md`](references/prompt-engineering-guide.md) |
| Worked examples (Claude / GPT / image) | [`references/examples/`](references/examples/) |
| Build a GPTs instruction | [`references/instructions/GPTs-Prompt-Generator.md`](references/instructions/GPTs-Prompt-Generator.md) |
| Build a Gems instruction | [`references/instructions/Gems-Prompt-Generator.md`](references/instructions/Gems-Prompt-Generator.md) |
| Build bot meta files (soul.md / CLAUDE.md / AGENTS.md) | [`references/instructions/Bot-Persona-Generator.md`](references/instructions/Bot-Persona-Generator.md) |

Procedure:

1. Classify the request (generate / fact-check / research / image / GPTs / Gems).
2. Read **only** the matched reference above.
3. For fact-check / research, the engineering guide's IFCN base template is
   mandatory — do not emit a generic XML prompt.
4. Produce the structured prompt; show the user the final prompt explicitly.

## Subcommands

This is a single-invocation skill — call `/prompt` with your task description (generate, review, refine, GPT/Gem instructions, image, fact-check, research). The skill routes internally to the matched reference (generator, engineering guide, examples, GPT/Gem instructions). Subcommands are:

| When to use | Call |
|---|---|
| Generate or refine a prompt | `/prompt ` — turns rough intent into a structured, framework-grounded prompt. |
| Review a prompt for robustness | `/prompt review: ` — checks completeness, phrasing, and edge cases. |
| Build GPT system instructions | `/prompt gpt: ` — generates structured GPT instructions using the GPT-Prompt-Generator framework. |
| Build Gem system instructions | `/prompt gem: ` — generates structured Gem instructions using the Gems-Prompt-Generator framework. |
| Build bot persona/meta files | `/prompt bot-persona:   ` — generates soul.md + CLAUDE.md/AGENTS.md using the Bot-Persona-Generator framework (used by guided onboarding). |
| Image generation prompting | `/prompt image: ` — applies the 5-stage image-generation framework (subject, style, composition, quality, mood). |
| Fact-check or research prompting | `/prompt research: ` — wraps the IFCN base template for verifiable claims and source-grounded research. |
| Study the framework | Load `references/prompt-engineering-guide.md` — deep techniques, mental models, and worked examples. |

## Attribution

- Upstream: `github.com/treylom/prompt-engineering-skills` (public, MIT) —
  see [`LICENSE`](LICENSE). Context-Engineering basis: Muratcan Koylan,
  *Agent-Skills-for-Context-Engineering*.
- Vendored here so the bot ships self-contained. The upstream maintainer's
  personal deploy script (`commands/prompt-sync.md`) is intentionally
  **excluded** (machine-specific, not relevant to distributees).

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [treylom](https://github.com/treylom)
- **Source:** [treylom/ThisCodex](https://github.com/treylom/ThisCodex)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-treylom-thiscodex-prompt
- Seller: https://agentstack.voostack.com/s/treylom
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.