OpenAI Skills (Official) — Prompts
Prompt 1: skill-installer — In-Session Discovery
File: skills/.system/skill-installer/SKILL.md
Technique: Self-referential bootstrap with graceful fallback
---
name: skill-installer
description: Install Codex skills into $CODEX_HOME/skills from a curated list or a GitHub repo path. Use when a user asks to list installable skills, install a curated skill, or install a skill from another repo (including private repos).
---
Install skills with the helper scripts.
## Communication
When listing skills, output approximately as follows:
"""
Skills from {repo}:
1. skill-1
2. skill-2 (already installed)
3. ...
Which ones would you like installed?
"""
After installing a skill, tell the user: "Restart Codex to pick up new skills."
Technique analysis: Prescribes the exact output format for skill listing (numbered list with installed annotations), making the agent's output predictable and parseable. The "Restart Codex" post-install message is baked into the skill instructions.
Prompt 2: openai-docs — MCP-Preferred Lookup with Fallback Chain
File: skills/.system/openai-docs/SKILL.md
Technique: Priority-ordered tool selection with explicit fallback chain
## Workflow
1. Clarify whether the request is general docs lookup, model selection, a model-string upgrade, prompt-upgrade guidance, or broader API/provider migration.
2. For model-selection or upgrade requests, prefer current remote docs over bundled references when the user asks for latest/current/default guidance.
- Use `mcp__openaiDeveloperDocs__search_openai_docs` to find the most relevant doc pages.
- Use `mcp__openaiDeveloperDocs__fetch_openai_doc` to pull exact sections.
- Use `mcp__openaiDeveloperDocs__list_openai_docs` only when you need to browse without a clear query.
## If MCP server is missing
If MCP tools fail or no OpenAI docs resources are available:
1. Run the install command yourself: `codex mcp add openaiDeveloperDocs --url https://developers.openai.com/mcp`
2. If it fails due to permissions/sandboxing, immediately retry with escalated permissions...
3. Only if the escalated attempt fails, ask the user to run the install command.
Technique analysis: Three-tier fallback chain: MCP first → escalated permissions retry → user manual install. The model-preservation instruction ("if the user names a target model like 'migrate to GPT-5.4', keep that requested target even if latest-model.md names a newer model") is a sophisticated anti-hallucination guard.
File: skills/.curated/gh-address-comments/SKILL.md
Technique: Number-and-select approval gate
## 2) Ask the user for clarification
- Number all the review threads and comments and provide a short summary of what would be required to apply a fix for it
- Ask the user which numbered comments should be addressed
## 3) If user chooses comments
- Apply fixes for the selected comments
Technique analysis: Forces agent to enumerate and summarize before acting. The "number all review threads" step converts an ambiguous PR into a discrete selectable set, preventing the agent from assuming which comments to fix.
Prompt 4: plugin-creator — Explicit Normalization Rules
File: skills/.system/plugin-creator/SKILL.md
Technique: Constraint specification with normalization rules
# Plugin Creator
## Quick Start
1. Run the scaffold script:
```bash
# Plugin names are normalized to lower-case hyphen-case and must be <= 64 chars.
# The generated folder and plugin.json name are always the same.
python3 .agents/skills/plugin-creator/scripts/create_basic_plugin.py <plugin-name>
**Technique analysis:** Embeds normalization rules directly in the instruction comment (`lower-case hyphen-case`, `<= 64 chars`) rather than in validation code, relying on the agent to self-enforce constraints.