Skip to content

Skillselionwith God's help

By phase

IdeaResearch & spot demand
ValidateScope & prototype
BuildFrontend, backend, MCP
ShipTest, review, deploy
LaunchDistribution, SEO, ASO
GrowAnalytics & retention
OperateMonitoring & uptime

By type

Skills
Marketplaces
MCP servers
Workflows

Skills Marketplaces MCP Learn Brief

Browse all tools →By phase

Idea Validate Build Ship Launch Grow Operate

Explore

Skills
Marketplaces
MCP
Learn
Brief

This week for builders

Five minutes, every Monday — the tools, releases and tactics for shipping solo.

Email address

unsubscribe anytime.

Skillselion

An independent, curated directory of AI-coding tools for solo builders - skills, MCP servers, marketplaces and workflows for Claude Code, Codex and Cursor, organized by the indie build journey from idea to operate.

ResourcesBrowse Skills Browse MCP Servers Browse Marketplaces Browse Workflows Best-of shortlists Glossary State of AI Skills 2026 Weekly Brief

CommunityAbout Learn Feedback Privacy Policy Terms Cookie Policy Accessibility Cookie settings

Last updated June 10, 2026 · © 2026 Skillselion. Skillselion is an independent project and is not affiliated with Anthropic, OpenAI, Cursor, Claude, Claude Code, or Codex.

Skills/NINGZIMU/CODEX PPT SKILL

ningzimu/codex-ppt-skill

1 skill · 988 installs · 1.2k stars · GitHub

Install

npx skills add https://github.com/ningzimu/codex-ppt-skill

Skills in this repo

1Codex PptCodex PPT is an agent skill for building slide presentations that depend on generated imagery. Before the first sample slide, the agent must read backend selection rules and decide between the environment’s built-in image generation tool—names like image_gen in Codex or image_generate in OpenClaw—and a local fallback driven by scripts/image_gen.py. Solo builders use it when they want deck visuals without manually juggling APIs: the skill prioritizes the native tool when callable, including gpt-image-2 style requests in Codex, and reserves CLI or OpenAI-compatible proxies for gaps or explicit user preference. Misclassification is a common failure mode, so the workflow stresses empirical availability checks rather than inferring capabilities from the agent brand. Output and temp directories are part of the repo hygiene model, keeping generated assets out of casual commits while supporting repeatable slide iteration.988installs