By popmechanic
Build applications where Claude Code CLI (claude -p) is the runtime — a server spawns Claude processes that read files, run commands, and return structured output, while a custom interface renders the results.
Build native desktop applications where Claude Code CLI (`claude -p`) is the runtime, using ElectroBun (Bun + system webview) with typed RPC — no HTTP server, no auth.
Build local web applications where Claude Code CLI (`claude -p`) is the runtime, using Bun (`Bun.serve()`) as a thin server that streams results to a browser. No auth, no deployment — the machine's existing Claude Code login is inherited.
Build server-based web applications where Claude Code CLI (`claude -p`) is the backend runtime — a Node/Express server spawns Claude processes to power a custom browser interface with streaming output.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
A Claude Code plugin for building applications where claude -p is the runtime.
Claude Code is usually an interface — you talk to it, it writes code. But claude -p is something more general: a process with access to the filesystem, the shell, and a broad set of tools, capable of returning structured output conforming to any schema you define. It's an engine you can embed inside an application of your own design.
Loom teaches Claude how to build on top of that engine. The architecture is simple. A server spawns a claude -p process with a prompt and a JSON schema, then streams events to a frontend via SSE. What the prompt asks, what shape the schema gives, and how the interface renders it — those are yours to decide.
This is the shift from Claude Code as a coding interface to Claude Code as a runtime. You inherit everything Anthropic built into that harness — tool use, context management, streaming, model selection, cost controls — and point it at whatever you want to make.
[Interface] ←→ [Server] ←→ [claude -p]
↕
[filesystem, shell, tools]
Three decisions define each app:
The server is a thin relay. Claude does the work. Your interface gives it form.
For a deeper walkthrough of the architecture, patterns, and design decisions, open the interactive explainer.
This plugin ships three skills — Claude picks the right one from your request:
claude -p, with per-user OAuth, rate limiting, and streaming to a browser UI.Bun.serve()) server that inherits your existing Claude Code login. No auth, no deployment.Loom is its own single-plugin marketplace. From inside Claude Code, add the marketplace, then install the plugin:
/plugin marketplace add popmechanic/loom
/plugin install loom@loom
loom@loom means the loom plugin from the loom marketplace. To apply it in the current session without restarting:
/reload-plugins
Verify and manage it with /plugin list, and pull future updates with /plugin marketplace update loom.
npx claudepluginhub popmechanic/loom --plugin loomVibes is for people who use Claude Code, but don't know how to code.
Julian project tools — deploy, manage, and interact with Julian instances
Autonomous, parallel, worktree-isolated execution of Superpowers plans via a committed Dynamic Workflow. Pairs with superpowers; invoke as /ultrapowers <plan-path> instead of subagent-driven-development.
Interactive plugin builder for Claude Code - serves as both an example plugin and a tool to create new plugins through guided prompts with specialized builder skills for each component type
LLM application development with RAG, embeddings, LangChain, and prompt engineering
Hot-reloadable versioned prompts with easy tools for prompt engineering, chain workflows, quality gates. Symbolic syntax: >>prompt --> >>chain @framework :: 'gate'
High-intelligence Claude Code copilot with deep code reasoning, evidence-driven planning, orchestration-first execution, model routing, context budgeting, CI/CD integration, enterprise security, plugin development, prompt engineering, performance profiling, agent teams, channels (event-driven autonomy with CI webhook, mobile approval relay, Discord bridge, and fakechat dev profile), interactive tutorials, LSP integration, security-hardened hook script library, MCP Prompts coverage, common workflow packs, runtime selection guide, computer-use patterns, checkpointing, scheduled-task blueprints, repo bootstrap scanner, hook policy engine (8 installable packs), layered memory deployment, role-based subagent packs (implementer, debugger, migration-lead, dependency-auditor, release-coordinator), 5 agent-team topology kits, autonomy operating mode (4 profiles + 3 gates), and a queryable 15-tool MCP documentation server with autonomy advisor.
CLI tool for running Claude Code SDK sessions with configurable modes
Real-time statusline HUD for Claude Code - displays context usage, tool activity, agent tracking, and todo progress