User-invoked codebase recon via an ant-colony pattern. The moderator carves a large or unfamiliar codebase into disjoint territories, fans out cheap Haiku scout subagents that explore in parallel and dump full findings to /tmp scratchpads (returning only terse pointers), then one Sonnet verifier cross-checks every…
A workflow for creating Agent Skills that work with tools such as Claude Code, Codex, and Cursor. It creates SKILL.md files, supporting references, trigger examples, validation checks, and installation files.
Index and route better-skills vault operations from the CLI. Trigger when users ask to manage better-skills or a skill vault (create, edit, delete, remove, search, import, clone, link, sync, backup, onboard, enable/disable vaults, inspect available vaults, move a skill across vaults). Do not use when asked to use a…
An issue-driven automation pipeline for GitHub repositories. It periodically checks issues, asks Claude Code to repair them, verifies the result later, and escalates cases that need a person.
Manage and rotate Claude Code accounts. Use whenever the user wants to switch/swap Claude accounts, check usage or how close they are to the 5h or weekly limit, jump to the account with the most headroom or the one that resets/frees up soonest, add/remove/rename a stored login, refresh an expired token, diagnose…
Bridge THIS Claude Code or Codex session to Telegram, Discord, Feishu/Lark, QQ, WeChat, or DingTalk so the user can chat with Claude from their phone. Use for: setting up, starting, stopping, or diagnosing the claude-to-im bridge daemon; forwarding Claude replies to a messaging app; any phrase like "claude-to-im"…
Create, modify, and debug agentic pipelines with Squid. Define multi-agent YAML workflows with spawn (OpenClaw, Claude Code, OpenCode), gates, parallel execution, loops, branching, restart loops, sub-pipelines, and structured approvals. Use when working with .yaml pipeline files or when the user mentions pipelines…
Design a focused Markdown loop.md contract for agent/automation loops: goal, termination, progress invariant, approval gates, measurement, residual routing, and optional subagent policy. Use for loop design only; output loop.md; never execute the loop, emit runtime code, or prescribe execution mechanics.
Write TypeScript code using the Thinkwell framework. Covers the @JSONSchema syntax, Plan fluent API, agent lifecycle, tools, skills, thought streams, and the thinkwell CLI.
Inject the experiment orchestrator role into the current session. Drives the full investigation pipeline: hypothesis refinement, environment exploration, experiment execution, transcript analysis, and reporting.
Get a second opinion from multiple AI models at once. Use this whenever you're stuck, want to validate an approach or design, pressure-test a plan or claim, or want diverse cross-model viewpoints on a hard decision - anything you'd reach for "peer review", "ask another model", or "council of models" for. Wraps the moa…
Autonomous full-stack software factory. Spec-driven. Builds a working React + Node app from a single idea — constitution file, persistent spec, task breakdown, tests before code, REST API, seed data, and polished UI. Enforces Anti-Generic UI (references/ANTIGENERICUI.md): no default analytics card-dashboard shells…
A planning framework for turning a vague request for repeated automation into a self-running workflow, called a loop. It defines the goal, triggers, task roles, connections, saved state, checks, and safety rules before producing commands for Claude Code or Codex.
Use Claude Fable 5 intelligence at close to Claude Haiku 4.5 cost. Three commands. distill has Fable write a reusable rulebook once. classify has Haiku apply it on volume, with deterministic escalation back to Fable. analyze has Fable judge a document corpus, with Haiku workers doing the reading when the corpus is…
Help developers use acp-cli — the Rust ACP CLI client for talking to coding agents (Claude, Codex, Gemini, etc.) over the Agent Client Protocol. Use when user asks about acp-cli commands, agent registry, session management, queue system, permission modes, or output formats.
Create tools for the Matimo SDK. Understand YAML tool definitions, execution types, authentication, parameter validation, and quality standards. Apply this skill when implementing Matimo tools.
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: