Building agents

38,215 mods in this category, of every kind an agent can take. Each one carries what it costs per session, what the scan found, and whether it is the original.

ralphmad

1730

hieutrtr/ralphmad

Plugin Claude Code

Bundles 4 commands, 1 hook · 29 tokens together

RalphMAD — Autonomous BMAD-METHOD workflow execution via ralph-loop. Templatized prompts that auto-populate from project context for reusable, project-agnostic SDLC automation.

not rated 39 6mo ago A tokens not measured original Apache-2.0

almogdepaz/wolfpack

Skill Claude CodeCodex

Create top-level Wolfpack sessions, spawn child agents, and inspect or control local/Tailscale sessions through the canonical CLI.

not rated 39 +1 changed 6d ago A SkillSpector: pass 31 tokens original MIT

meta-skills

1734

CheshireMew/meta-skills

Skill Claude CodeCodex

A guide for designing, reviewing, improving, testing, and distributing skills for coding agents.

not rated 39 7d ago A 174 tokens original MPL-2.0

kimi

1735

linxule/kimi-plugin-cc

Plugin Claude Code

Bundles 12 skills, 11 commands, 7 agents, 1 hook · 1,124 tokens together

Claude Code plugin that delegates read-only review, challenge, free-form ask, and write-capable rescue work to the local kimi-code CLI. Includes an opt-in stop-time review gate and a PreToolUse hook that enforces per-command safety.

not rated 39 +1 changed 4d ago A tokens not measured original Apache-2.0

singular-orchestration

1736

alex-reysa/singular-lite

Skill Claude CodeCodex

Operate Singular (repo "singular-lite"; CLI command singular) — an autonomous multi-agent orchestration engine that drives parallel AI coding agents against a git repo through a three-tier L0/L1/L2 model with DAG plans, durable leases, gate pipelines, and worktree isolation. Use whenever the user wants to orchestrate…

not rated 39 +1 9d ago A 185 tokens GPL-3.0

crewai-skills

1737

crewAIInc/skills

Plugin Claude Code

Bundles 3 skills · 242 tokens together

CrewAI development skills for Claude Code covering agent design, task configuration, structured output, and debugging common pitfalls.

not rated 39 3mo ago A tokens not measured

smart-mcp-proxy

1738

Dumbris/mcpproxy

MCP server Claude CodeCodexCursor +2

Smart MCP Proxy - A federating gateway for MCP servers with intelligent tool discovery. Runs locally from the smart-mcp-proxy Python package.

not rated 39 1y ago A tokens not measured original MIT

local-skills-mcp-guide

1739

kdpa-llc/local-skills-mcp

Skill Claude CodeCodex

Repository implementation guide for the local-skills-mcp codebase. Use when asked: how src/index.ts and src/skill-loader.ts work together; where MCP tool handlers are defined and registered; how getAllSkillsDirectories priority and override behavior works; how local-skills-mcp discovers skills and merges metadata…

not rated 39 yesterday A SkillSpector: pass 98 tokens original MIT

curator-evolution

1741

pingchesu/hermes-curator-evolver

Skill Claude CodeCodex

Use when interpreting Hermes Curator Evolver evidence reports, historical session backfill, proposals, verifier results, candidate search, auto-run output, timer installs, or guarded apply manifests.

not rated 39 +1 changed 10d ago A SkillSpector: warn 41 tokens original MIT

agentskillexchange/skills

Skill Claude CodeCodex

5-phase repeatable structure for autonomous agent sessions: context-load, tiered work-selection, coordination claim, execute, and persist-learning. Prevents duplicate work across concurrent sessions and ensures every run produces durable artifacts. Runtime-agnostic — Claude Code, gptme, Codex, or any…

not rated 39 +1 today A SkillSpector: warn 69 tokens original MIT

skillless

1743

0oooooooo0/skillless

Plugin Claude Code

Bundles 3 skills, 2 commands · 0 tokens together

Auto-discovery plugin for Claude Code skills - automatically recommends and installs relevant capabilities without users needing to know about skills.

not rated 38 6mo ago A tokens not measured original MIT

think-tool

1744

cgize/claude-mcp-think-tool

MCP server Claude CodeCodexCursor +2

MCP Think Tool server for Claude Desktop. Runs locally from the @cgize/mcp-think-tool npm package.

not rated 38 1y ago A tokens not measured original MIT

skill-manager

1745

egebese/skill-manager

Skill Claude CodeCodex

Detect project tech stack and disable irrelevant skills to save context window space. Use when the user says "manage skills", "optimize skills", "disable irrelevant skills", "skill manager", or asks which skills are relevant for this project.

not rated 38 6mo ago B 50 tokens original MIT

fewword

1746

sheeki03/Few-Word

Plugin Claude Code

Bundles 1 skill, 21 commands, 4 hooks · 257 tokens together

Context engineering plugin - tiered offloading, smart summaries, secret redaction, TOML/JSON config, command aliases, MCP allowlists, and self-diagnostics.

not rated 38 7mo ago A tokens not measured

code-pilot

1747

ysicing/code-pilot

Plugin Claude Code

Bundles 1 skill, 10 commands, 5 agents · 349 tokens together

A development workflow system with more than 15 commands and specialized AI roles for handling software tasks. The available description identifies it as an intelligent development-automation tool, but does not explain each command or role.

not rated 38 5mo ago A tokens not measured original Apache-2.0

ai-native-code

1748

tianji-qingtian/AI-Native-Code

Skill Claude CodeCodex

A software-development working method for using Claude Code as the main coding tool, for both new projects and existing codebases.

not rated 38 4mo ago A 280 tokens

modelharness

1749

vitaliikapliuk/modelharness

Plugin Claude Code

Bundles 3 skills, 3 commands, 1 agent, 1 hook · 234 tokens together

Fable 5 behavioral patterns as a harness for Opus 4.8: grounded progress, self-verification loops, delegation triggers, file-based memory, autonomy calibration. Zero-config: installs a SessionStart hook; skills auto-trigger.

not rated 38 3mo ago A tokens not measured original MIT

regenrek/codex-proofloop

Skill Codex

Coordinate Planr-backed or project-configured Herdr workflows in which Sol owns intake, architecture, design-sensitive implementation, integration, and the final decision; Luna Max serves only as an explicitly selected read-only silent sentinel, fresh bounded verifier, or interactive operator; and Fable supplies at…

not rated 38 1mo ago A SkillSpector: pass 120 tokens original MIT

HAL

1752

DeanWard/HAL

MCP server Claude CodeCodexCursor +2

HAL (HTTP API Layer) - An MCP server that provides HTTP API capabilities to Large Language Models. Runs locally from the hal-mcp npm package.

not rated 38 10mo ago A tokens not measured original MIT

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: