3amigos

A coordinator for coding runs, with a time box only when the owner names one, shared as a team plugin — one main session that coordinates background subagents in git worktrees and does the light work (reads, git inspection, notes) itself, with Codex reviews first (Opus when Codex is not installed), a 15-minute health check, and destructive or outward actions queued for the user at the end of the run. Run /captain:captain-setup once first to choose notifications and quiet hours. Code and functionality only; not for production, devops or external writes beyond listed dev targets.

2 skills · 4 agents · hooks

Mentorano's fork of kenryu42/cc-safety-net: a PreToolUse hook that blocks destructive commands and secret file access. The fork lets a heredoc written through `command cat` stay text, lets read-only git commands name the policy file, and reads the project policy from the repository root when the agent works in a subdirectory or worktree.

5 skills · hooks

Clean Code and Clean Architecture rules, source measurement, and logging for Python and TypeScript. Includes architecture-fundamentals for system decisions and reviews: business drivers, data ownership, runtime coupling, contextual style trade-offs, risks, ADRs and fitness functions.

5 skills

Committed project memory for Claude Code — one topic per markdown file in .agent-memory/ at the repo root, loaded natively as rules through a committed .claude/rules/project-memory symlink (path-scoped files load when a matching file is touched), with a write trigger every session and subagent reads, a file-state health check at session start, and a defragment skill that merges, splits and trims the memory in one reviewable diff. The plugin never commits.

2 skills · hooks

Small helpers for working with Claude Code — start a new background Claude Code session with Remote Control in the same repository and get its link, to continue from a phone or the web. Also a one-by-one pass over review findings that closes every decision before the final report.

3 skills

Build a live status page for a repository's recent branches — what exists, who moved it, how far it got, what it touches, and what its own plan documents say the goal is. Every number is derived from git and gh, never remembered.

1 skill

Deep vulnerability scanning of your own code, run entirely inside your Claude Code session at a chosen effort tier, with every finding challenged before it is reported and the verification tally computed in code. Turns surviving findings into targeted patches, each verified by a panel of agents, that you apply when you choose.

1 skill · 9 agents · hooks

Independent auditing for AI agents, open source. Claude discovers your setup, confirms the cost, and audits your agent against the one question other tools skip: is it doing the job it's supposed to do, given your business rules and org structure? Test any model (Anthropic, OpenAI, Gemini, Azure, Bedrock, …) graded by self, one independent judge, or a cross-vendor panel; each provider is billed to its own account via a key in your settings, never the command line. Results render as an interactive artifact (with a static-report fallback).

1 skill · hooks

Production incident management, triage workflows, and automated incident resolution

4 skills · 2 commands · 6 agents

Speed up contract review, NDA triage, and compliance workflows for in-house legal teams. Draft legal briefs, organize precedent research, and manage institutional knowledge.

9 skills · 7 MCP servers

Email-workflow skill pair: inbox-setup builds your taxonomy/KB; inbox-triage classifies + drafts (drafts-only, never auto-send). KB-file contract between them. Path-B from megaprompts 06+07.

2 skills · 2 commands · 2 agents

Brainstorm, plan, debug, review, and compound learnings with AI agents

39 skills