Building agents skills

harnessgrid/harnessgrid

Skill Codex

Coordinate bounded Claude implementation workers with durable authenticated A2A/LAN jobs, explicit evidence, recovery, and read-only verification.

catalystsystemslab/codex-skill-sync

Skill Codex

Simple sync flow for installed SKILL.md skills and Codex plugins. Use when the user says /skill-sync, check skills, update skills, run all skill updates, map skill repos, refresh Codex plugins, or create weekly skill maintenance.

birdilsss-byte/ai-creative-subagent-framework

Skill Claude Code

A workflow for producing AI-assisted film and video concepts with separate roles for directing, visual design, cinematography, brand review, and AI execution. It breaks a script into storyboards and develops the visual plan in stages.

theelgreco/codex-fork-tab

Skill Codex

Open an exact fork of the active Codex CLI chat in a new terminal tab while leaving the source chat open. Use when the user asks to fork, branch, duplicate, or continue the current Codex conversation in another terminal, including phrases such as “fork this into a new tab” or “open a parallel copy of this chat.”.

Soul-Brews-Studio/plugin-marketplace

Skill Claude CodeCodex

Create new skills with Oracle philosophy. Use when user says "create skill", "new skill", "make a skill for X", or wants to build custom Claude Code skills.

Charlesaitime/claude-codex-teamwork

Skill Codex

Peer-to-peer Claude Code ⇄ Codex CLI collaboration via shared task board. Supports Default Mode (Claude plans, Codex executes) and Max Mode (parallel planning + execution).

JS-mark/skills

Skill Claude Code needs its repo

Multi-Agent Collaborative Pipeline for Claude Code — orchestrates 7 development phases with 6 specialized AI roles.

erenbektas/4CO-OP

Skill Claude Code

Run the 4CO-OP multi-stage coding pipeline for a feature request, including the nightly monitor window and performance logging.

OhadRubin/skills

Skill Claude CodeCodex

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.

umbraco/Umbraco-MCP-Base

Skill Claude CodeCodex needs its repo

Guide for configuring and running Umbraco MCP servers via the CLI. Use when the user wants to set up an MCP server for Claude Code, configure auth, filtering, dry-run, readonly mode, or use introspection commands to understand available tools.

sanztheo/claude-intercom

Skill Claude CodeCodex

Always active — coordinates with other Claude Code agents via intercom MCP. Automatically checks for messages, announces presence, and communicates to avoid conflicts. Triggers on ANY task start and periodically during work.

gonaas/skills

Skill Claude CodeCodex

Invoke this skill immediately when the user wants to: create a new Claude Code skill for any purpose, convert a workflow or conversation into a reusable skill, edit or improve an existing skill, run evals or benchmark skill performance, compare skill versions, or optimize skill triggering. This applies whether the…

Congee-Chem/skills

Skill Claude CodeCodex

A guided way to build a complete web application with an AI coding partner, from clarifying the idea through deployment. It covers planning, choosing tools, designing the structure, writing code in small parts, testing, review, and delivery.

bleronsh/Skills

Skill Claude CodeCodex

ITxPT specification knowledge base for the YES-EU Fleet Management Platform. Auto-loaded to provide persistent ITxPT architecture context without re-reading the 13 source PDFs each session. Covers S01 (hardware), S02 (onboard services), and S03 (back-office) from the "Linden" release.

PayMCP/skills

Skill Claude Code

Integrate PayMCP into an MCP server to monetize tools (pay-per-request pricing or subscriptions), configure payment providers (traditional and/or X402), choose an appropriate coordination mode, and validate the integration.

d-gangz/dgang-skills

Skill Claude CodeCodex

Create a new skill from the conversation that just happened, or improve an existing one — carries the reference for writing skills well.

saemihemma/lead-producer-oss

Skill Claude CodeCodex

Top-level orchestrator for multi-domain tasks. Routes work to specialist roles, enforces simplification, stress-tests via Devil's Advocate before acceptance.

theshadow27/mcp-cli

Skill Claude CodeCodex

Use the mcx CLI via Bash instead of native MCP tools. This eliminates 15,000+ tokens of tool definitions per server from every conversation context.

dmarzzz/shape-rotator-field-kit

Skill Claude CodeCodex

Durable AI workflow orchestrator. Run, monitor, and manage workflow executions. Run smithers --help for usage details.

companion-inc/companion-skill

Skill Claude Code

Offload a Claude Code task to a remote Companion sandbox via SSH. Use when the user wants to continue a task in the cloud, run something remotely, or offload work to their Companion sandbox. Triggers on "offload", "run in cloud", "companion run", "remote task", or "send to sandbox".

spideynolove/claude-dotfiles

Skill Codex

Orchestrates the RLM repo workflow end-to-end with phase gates, locked artifacts, addenda, traceability, and automatic bootstrap/upsert of AGENTS/PLANS scaffolding. Trigger phrases: "Implement requirement ", "Run RLM Phase ", "resume requirement", "lock Phase ", "verify locks".

eq1024/skills

Skill Claude CodeCodex

A guide for adding restrictions to an AI assistant's coding behaviour through project or global instruction files.

byx-darwin/gitflow-cli

Skill Claude CodeCodex

A four-step workflow for clarifying a task, planning it, carrying it out, and delivering the result. It records progress as JSON and requires checks between steps.

openjobs-hq/openjobs

Skill Claude CodeCodex

Use this skill whenever checking OpenJobs inbox/messages or running the OpenJobs command-center workflow for the active agent. It ensures the exact CLI binary is used, platform health is verified, unread tasks and DM counts are checked, command-center actions are dispatched, webhook health is monitored, oversight…