Building agents commands

2,458 tagged Building agents, measured the same way as everything else here.

Browse within: claude-code-plugin 217cli 154ai-coding 126agentic-ai 125agents 123LLM 118Multi-Agent 70copilot 66agent 64agentic-workflow 64ai-tools 57skills 55Autonomous Agents 48Orchestration 46

skill-csk

97

byerlikaya/claude-starter-kit

Command Claude Code needs its repo

Author or change a kit component (skill / agent / command) against the template and its gates.

not rated 22 6d ago A 24 tokens original MIT

agent-harness-setup

98

charlie947/agent-harness-starter

Command Claude Code

Set up your agent harness — guided walkthrough that builds your /.claude/ structure (Personalisation, Memory, Skills, Agents) or a project CLAUDE.md.

not rated 22 3mo ago A 34 tokens original MIT

openstation.check

100

leonprou/openstation-legacy

Command Claude Code

Check on a running or completed agent session. $ARGUMENTS = task ID or run ID. Analyzes logs for health, progress, issues, and blockers.

not rated 22 2mo ago A 38 tokens

handoff

101

legoambarish/portable-handoff

Command Claude Code

Use the active Claude Code model's current session as the semantic source. Preserve corrections, constraints, decisions, verification truth, and one exact next action. When shell access is available.

not rated 21 17d ago A 0 tokens original Apache-2.0

panquequelol/coding-agents

Command Claude Code

Solve the user task with one Oracle plan, at most one Oracle finding review, and a strict Sentinel approval gate. The main agent owns all edits and checks. Oracle and Sentinel do not edit files.

not rated 21 yesterday A 0 tokens

proofline

103

giang6283623/minimal-vibe-coding-kit

Command Claude Code

Activate Proofline governance with bounded roles, independent challenge, typed signals, and evidence-bound handback.

not rated 21 +3 9d ago A 20 tokens original MIT

swarm-init

105

natea/ExoMind

Command Claude Code

Initialize a new swarm with specified topology.

not rated 21 10mo ago A 0 tokens original MIT

create-command

106

Saik0s/agent-loop

Command Claude Code

Guide for creating new custom Claude commands with proper structure.

not rated 19 1y ago A 11 tokens

add-skill

107

2ykwang/agent-skills

Command Claude Code

Scaffold a new skill with validation, plugin generation, and README update.

not rated 19 +1 1mo ago A 15 tokens original MIT

forge

108

zakelfassi/skills-driven-development

Command Claude Code

Forge a new SkDD skill from a repeated pattern (shortcut for the skillforge meta-skill).

not rated 18 1mo ago A 20 tokens original MIT

hm-night

109

JavanC/Homunculus

Command Claude Code

Run the evolution pipeline. Depth depends on evolution-config.yaml (tier + schedule).

not rated 15 3mo ago A 0 tokens original MIT

scan-project

111

wonsukchoi/domain-experts

Command Claude Code

Scan a project and suggest domain-expert skills it would benefit from, then draft the ones the user picks.

not rated 15 +1 7d ago A 22 tokens original MIT

squad-manager

113

rcelebrone/DotAgents

Command Claude Code

Protocolo central que rege o trabalho da squad multi-agente DotAgents. Sempre ativo.

not rated 14 29d ago A 23 tokens

set-hermes-telegram

114

chankov/agent-fleet

Command Claude Code

Install/status the Hermes liaison or turn its Telegram askuser bridge on/off.

not rated 14 +1 today A 14 tokens original MIT

hub

115

shahtuyakov/claude-setup

Command Claude Code

Process delegation requests from agents and orchestrate multi-agent workflows.

not rated 13 8mo ago A 0 tokens original MIT

cos

117

George-RD/claude-next-level

Command Claude Code

Parse a work description, analyze dependencies, route to workflows, dispatch agents in waves.

not rated 12 2mo ago A 19 tokens

mcp-connect

118

algolia/skills

Command Claude Code

Configure a supported MCP client with the Algolia MCP server.

not rated 12 4d ago A 12 tokens original MIT

gedd-chat

119

aws-samples/sample-GEDD

Command Claude Code ✓ vendor

You are a GEDD coaching assistant. You guide the user through building a golden evaluation dataset for their AI agent using Open Coding methodology, then help them evaluate and annotate responses — all without leaving Claude Code.

not rated 12 1mo ago A 0 tokens MIT-0

speckit.assess.decide

120

github/spec-kit-copilot

Command Claude Code ✓ vendor

Command "speckit.assess.decide" from github/spec-kit-copilot, covering when to run it, how to ask — always use the askuser tool, after the answers and what to ask about at the decide stage.

not rated 12 2d ago A 0 tokens 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: