Development plugins

Teach an agent to write correct onesie invocations.

6 skills · hooks

Codebase-aware specification engine. Runs a grounded interview that captures implicit facts as A-AUTO-NNN entries, emits typed Global Invariants / State Transitions / Contracts tables, runs an adversarial spec review at R3.5, and writes a foundry-ready spec with version frontmatter — locked, classified, and verifiable. Brownfield runs produce a flow delta against the existing system; greenfield runs produce an end-state spec; cosmetic runs skip flow mapping. Forge plans. Foundry builds.

4 commands · 3 agents

Language servers for Claude Code that stay in step with the disk and stay quiet: Dart, PHP, TypeScript and JavaScript, Python, Go, Swift and HTML behind one proxy that holds retracted diagnostics, drops ghosts of deleted worktrees, filters hints and environment noise, and stops idle servers.

LSP

Verdicts an agent loop can trust: skills that make Claude actually run, test and verify a Godot 4 project instead of trusting exit codes.

6 skills · 1 agent

Create, build, check, install and share Localix mini apps from Claude Code or Localix.

5 skills · 1 agent · hooks

Create, build, check, install and share Localix mini apps from Claude Code or Localix.

5 skills · 1 agent · hooks

Agent Rally Point — repo-local coordination surface for parallel coding agents. Claude Code has a plugin integration; Codex has Rally skills and needs project hook configuration; Cursor uses project hook configuration; Gemini and other hosts use the CLI loop. Records claims, checks write boundaries, routes handoffs, and exposes room state via the rally CLI. Skills: agent-rally-point (per-session loop), rally-workflows (fan-out + dynamic-workflows protocol), mini-loop (per-task quality gate). Hooks auto-fire only in supported, configured projects on SessionStart + PreToolUse to surface room awareness and deconflict before edits; advisory by default (the edit goes through), with opt-in strict and fail-closed switches that block. Backend: rally CLI (cargo install --path crates/rally-cli).

4 skills · 1 command · hooks

Opinionated defaults and scaffolding for rapid MVP static websites using Next.js or 11ty monorepo patterns

5 skills · 1 command

Multi-agent code review for pending Perforce changelists (/p4-code-review) -- auto-shelves, reviews, cleans up.

1 skill · 5 agents

Architecture intelligence: sizing, design, review, and ADRs — proportional to your actual scale

5 skills · hooks

Feature-driven development assistant for OpenSaaS Stack. Build complete applications by describing what you want to build instead of configuring infrastructure.

1 skill · 2 commands · 2 agents · 1 MCP server

Team hub over MCP: project memories, plans, skills and knowledge graphs, with Claude Code's memory files and plan copies kept in step by hooks.

4 skills · 1 MCP server · hooks

Sends opt-in, sanitized Claude Code workflow and conversation telemetry to SkillBench for organization-level developer skill analytics.

4 skills · hooks

Generate PDFs from LaTeX templates. CLI, Go library embedding, template syntax, and live preview over SSE/WebSocket/HTTP2.

5 skills

squidapp (Next.js + NestJS/Prisma + Postgres scaffolder) + squidops (toolchain doctor), plus squidup/squidupgrade.

5 skills

Interview-driven project scaffolder for Claude Code. Routes new projects to the right Spec-Driven Development methodology (Superpowers, Spec-Kit, MUSUBI, OpenSpec, or lean) based on Piskala's three-level rigor taxonomy and Wasowski's four decision contexts.

3 skills · 2 commands

Parallel AI engineering control-plane plugin with task-graph scheduling, file ownership isolation, cost-aware model routing, 9-gate system, RBAC governance, and audit trail

4 skills · hooks

Shared Kubernetes controller skills for developers contributing to any Envoy control plane — reconcile loops, finalizers, controller testing, leader election, and CRD lifecycle

5 skills

Translate a validated AGER/OKF AgentGraph into a Claude Code plugin (skills, agents, commands, hooks).

3 skills · 1 command · 1 agent

Skills for prose written for other developers: the shared voice, code comments, commit messages, cover letters, and version changelogs

5 skills

Use when writing, reviewing, or debugging Terraform HCL that uses any Zscaler provider (zpa, zia, ztc, zcc), or when designing the structure / CI/CD / state organization / security pipeline of a Zscaler-Terraform repository. Routes to a per-product skill (zpa-skill, zia-skill, ztc-skill, zcc-skill) for resource catalog, auth, and lifecycle quirks; routes to best-practices-skill for cross-cutting engineering discipline (state, CI/CD, secrets, testing, modules).

5 skills

Process-oriented software development workflow: plan, build, test, debug, review, and ship

4 skills · hooks

The outer loop. Drives one turn of design, the sayt TDD loop, and review, routing each obligation to the budget that can settle it.

2 skills · 2 commands · 1 agent

Delphi/Object Pascal: Language Server (Diagnosen, Navigation) und MCP-Werkzeuge von Atlas

2 skills · 1 MCP server · hooks · LSP

Sends opt-in, sanitized Claude Code workflow and conversation telemetry to SkillBench for organization-level developer skill analytics.

4 skills · hooks

Build and deploy internal web apps on Backyard: the Backyard skill, /backyard:new, /backyard:deploy and /backyard:docs, and the backyard MCP server.

1 skill · 3 commands · 1 MCP server

Evidence-first coding agent with adversarial multi-model review, SQL verification ledger, and automatic rollback.

2 commands · 1 agent · 1 MCP server · hooks

Set up, audit and work within tool-enforced guardrails for coding agents in TypeScript repos.

5 skills

Starter specialist agents: architect, backend, frontend, design and devops.

5 agents

Lean spec-to-code build workflow: code-driven preflight and planning, builders that reuse what the project already has, each slice spec-checked as it lands, the project's real end-to-end check run in code, and at most two batched fix rounds. A separate plugin; forge and foundry are unchanged.

2 commands · 3 agents