Skills, prompts, AGENTS.md files and MCP servers that Codex reads. Skills are shared with Claude Code; the rest is Codex-native. 282,341 indexed, measured the same four ways as everything else here.
Produce a publishable piano score and matching 1080p60 falling-notes video from a user-provided audio recording, including metadata research, cover art, PDF/MusicXML/MIDI exports, review-gated YouTube publication, and mymusic5 sheet-music listing. Use for piano transcription, sheet-music delivery, falling-notes video…
Create and edit a product, launch, demo, or social video from the current project with 21st video MCP tools. Open the editable project alongside the chat, apply small revision-aware edits, check rendered frames, offer visual variants, and export the result. Use when the user asks for a video, says "make a launch video…
An auditing and organizing skill for an AI project's instructions, rules, skills, and supporting context. It finds duplication, conflicts, outdated information, and prompt injection, then creates reviewable, reversible change proposals or applies approved changes.
Manage persistent wiki via the kb CLI or HTTP API. Use when storing, retrieving, updating, or searching knowledge that should persist across conversations. Triggers on remember, store, recall, look up, find in memory, organize knowledge.
Guide for creating content on ayokoding-web, a Next.js 16 fullstack content platform (ayokoding-web). Covers bilingual content strategy (default English), tRPC API, content workflow, and ayokoding-web specific patterns. Essential for content creation tasks on ayokoding-web.
Test mobile apps with MobiLoop MCP, collect device evidence, and fix and retest confirmed failures within the user's scope. Use for Flutter, React Native, Android, or iOS QA on an emulator, simulator, or device, including relevant mobile security checks.
Library of 670 proven prompt templates (1,007 originals, variants merged) for software work. Use whenever the user asks for "a prompt for X", wants a better prompt, or starts a coding, frontend, backend, full-stack, API, database, DevOps/CI/CD, cloud, AI/ML/LLM/RAG, debugging, code review, architecture/system design…
A command-line guide for managing MCP Apps in mcpcat. MCP Apps are hosted application resources that can include tools and other app content for connected AI systems.
Email infrastructure and authentication for developers: domains, inbox routes, auth apps, deploys. Runs locally from the @smbcloud/cli npm package. Needs 1 environment variable to run.
Use when reviewing an agent tool that reads or writes PostgreSQL, such as whether an agent could read another customer's rows, lose a user's requirements on a retry, or write the same business effect twice. Not for building a new agent.
Persistent memory for AI agents — recall past decisions and record new ones across sessions, search a public research knowledge graph, and read the Open Deal Board (awareness.market).
Create or edit Echo slash commands for prompt text, standalone plugin tools, or app views and features. Includes every command option, registration examples, and how to generate a backing plugin when needed.
Open agents or browser tabs in existing TerminalVibe splits or new splits, navigate browser panes, arrange panes, inspect terminal output and browser content, and send follow-up input. Use when the user asks to open, inspect, or coordinate panes, agents, terminals, or embedded browser tabs inside TerminalVibe.
Use this skill when writing, rewriting, or critiquing LinkedIn cold outbound messages, connection requests, DM icebreaker sequences, or follow-ups. Encodes John Peslar's cold outbound playbook, filled out with Kellen Casebeer / GTM Cafe doctrine and top-engagement practitioner research from YouTube and Reddit.…
How to write and run a dynamic workflow script with the runworkflow tool: a Python script that orchestrates many agents with agent, parallel, pipeline, and phase. Use when a task splits into many independent or staged pieces of agent work (fan out over files, modules, bugs, or review angles, then aggregate), or when…
Close the Gumroad books on the self-hosted double-entry ledger. Use for the daily close (feeds synced, balances tie, nothing unposted) and the month close (journal entries, reconciliation to bank and processors, review markers written). Replaces the retired QuickBooks workflow.
How to call the Parism MCP tools (mcpparismrun, queryresult, describe) with the fewest calls and tokens. Load this BEFORE your first mcpparism call in a session, whenever those tools are available. Covers one-line commands with pipes, picking fields, filtering, sorting and counting rows on the server, reading large…
Discover plausible inputs, state transitions, orderings, and dependency behaviors that violate an expected contract, prove the smallest genuine surprise, and forge it into a deterministic regression test. Use for edge-case design, negative testing, failure injection, fragile happy paths, or bugs that need a durable…
Inventory, classify, route, catalog, relocate, or validate Codex skills and generate a skills menu. Use for skill-library organization or when Codex must find task-relevant skills. Classification, relocation, creation, revision, and promotion require user approval.
Interpret and act on a moire finding. Use when moire check or moire verify reports a CONFLICT or BROKEN result against another agent's worktree, or when deciding how to respond to an overlap with concurrent in-flight work.
Use when designing, creating, modifying, packaging, installing, reviewing, auditing, or diagnosing a DeepSeek Harness plugin: a live dynamic Cordis plugin driven by cordisdefine, a source-backed workspace plugin under packages, or an out-of-tree installable bundle using dsh.bundle and cordis.patch.yml. Also use for…
★not rated 11 3mo agoA130 tokens
originalMIT
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: