38,494 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.
A tiny phase gate for coding agents: your shell checks decide pass/fail (never the LLM), your workflow routes the agent (advance / retry / escalate / done). The agent drives, headsign holds the state.
★not rated 1
changed 4d agoA
tokens not measured
originalMIT
An agent driving a browser tends to start a new one rather than use the one already open, and to size its work off the core count. This warns once per session when a tool payload is about to do either. Never blocks.
★not rated 1 1mo agoA
tokens not measured
originalMIT
Agentic test automation guidelines based on the CIGE standard (Context, Intent, Guardrails, Execution) for authoring self-healing, goal-driven AI test cases.
★not rated 1
changed 9d agoA
tokens not measured
originalApache-2.0
Rigorous, verifiable agent workflows for Claude Code. A router over playbooks, an opt-in spec path, durable state on disk, and gates that are code rather than prose.
★not rated 1 27d agoA
tokens not measured
originalMIT
A plugin marketplace package built around AI-driven domain-driven design, a way to organise software around the business concepts it models. Its description lists 42 skills, 11 agents, and hooks, including a default secret-protection hook.
The Superpowers software-development methodology executed as a guarded state machine: 21 typed nodes, 96 total exit guards, 9 bounded retry loops, durable run state that survives compaction, and a ledger recording every step that could not run.
★not rated 1 1mo agoA
tokens not measured
originalMIT
Audit how well a project takes advantage of Claude Code's extensions (CLAUDE.md, settings.json, skills, subagents, rules, MCP) and produce a 0-100 score with recommendations.
Creates custom auto-develop loops for any feature in any codebase — analyzes features, builds capability specs, and generates autonomous iterative development skills that add new functionality.
★not rated 1 4mo agoA
tokens not measured
originalMIT
Approval gate for instruction files. Surfaces new or modified CLAUDE.md, .claude/rules/ files (project and user-global) and the files they @import at session start; the user explicitly affirms them before Claude treats them as trusted.
★not rated 1
changed yesterdayA
tokens not measured
originalMIT
Advisor Strategy skill for Claude Code — spawns a stronger model (Opus) as a strategic advisor for architecture, security, debugging, and performance decisions.
When a convention bites you, fix the convention — an evidence-gated repair loop (detect → classify → verify → codify → propagate → record) plus budgets, demotion and archiving that keep rule docs from growing without bound. Stack-agnostic.
★not rated 1
changed 13d agoA
tokens not measured
originalMIT
The full Cantos system as a one-command install: a self-improving, multi-agent assistant orchestrator for Claude Code. /cantos:init scaffolds the whole writable system into a project; /cantos:start runs it for one session. Bundles the orchestrator, specialist assistants, sub-agents, workflows, references, and a…
★not rated 1 todayA
tokens not measured
originalMIT
Skills and agents for Claude Code — visual verification, shareable HTML artifacts, arXiv research intake, model routing, deck and theme systems, and specialist advisor agents.
★not rated 1 21d agoA
tokens not measured
originalMIT
Explore earning with AI through task trading and Skill commerce: find paid work, sell Agent Skills, deliver results, and buy or run Skills. Check live opportunities, fees and Circle Credits terms. No guaranteed income.
★not rated 1
changed 13d agoA
tokens not measured
originalMIT
Turns Claude Code into a disciplined engineering team: it frames the goal, plans behind an approval gate, routes each step to the best specialist plugin (with built-in fallbacks), and runs a bundled Test Architect, an adversarial Security Gate, a two-axis code review, and an ownership review that makes sure a…
★not rated 1 2mo agoA
tokens not measured
originalMIT
Spawn separate zero-token Claude Code sessions against a local model (llama-swap / llama.cpp) for long-running or simple side tasks — keep your main session in its prime context.
★not rated 1 1mo agoA
tokens not measured
originalMIT
Minimal Exolvra Genesis loop orchestrator for Claude Code: a lead agent iterates builder subagents against blind, fresh-context critics comparing real output to a concrete quality bar, until the assembled work wins twice in a row.
★not rated 1
changed yesterdayA
tokens not measured
originalMIT
Delegate work packets from a lead Claude Code session to executor sessions in their own iTerm tabs, with session reuse instead of always spawning fresh.
★not rated 1
changed todayA
tokens not measured
originalMIT
Build Claude Code skills, agents, hooks, and MCP servers from plain-language descriptions. Generate, validate, and ship plugins without writing boilerplate.
★not rated 1 4mo agoA
tokens not measured
originalMIT
Your AI dev brigade for Claude Code. A chef (sprint planner) coordinates specialist agents (frontend, designer, nodejs, golang, devops) through wave-based execution with fan-out review, Bug Council for hard bugs, and sprint learnings between iterations. Zero dependencies.
The team-architecture factory for Claude Code — a meta-skill that turns a domain description into an agent team and the skills they use. Three native execution modes (Workflow orchestration, persistent named-agent collaboration, sub-agent delegation), six team-architecture patterns, and workflow-native quality…
★not rated 1 23d agoA
tokens not measured
originalApache-2.0