perfup
01Command Claude Code
Autonomous performance optimization: research, PoC, benchmark, implement, review, PR.
49 tagged local llm, measured the same way as everything else here.
Browse within: gemini-cli 16agentic-memory 15gdpr 15ollama 13accessibility 8kokoro 8llama-cpp 7english 5offline 5study 5
Command Claude Code
Autonomous performance optimization: research, PoC, benchmark, implement, review, PR.
Command Claude Code
Translate the following documentation files to all supported languages.
Command Claude Code
Wire Cursor, Claude Code, Codex, or Honcho to the local netllm router.
Command Claude Code
Troubleshoot netllm misconfigurations (PATH, providers, mDNS, agent health).
Command Claude Code
Configure multi-machine LAN mesh for netllm (mDNS, peers, gateway).
Command
Delete a memory permanently. Asks for confirmation first.
Command
Health check — package version, installed payload, chatlog DB row count, per-agent hook state.
Command
List all m3-memory slash commands and what they do.
Command Claude Code
Review a pull request with one agent for small changes and a full agent set for large changes.
Command Claude Code
Hands the P1–P5 findings from the most recent /code-review to Hermit so Claude doesn't burn tokens re-reading every file and retyping every edit. Claude stays on judgment; Hermit does the mechanical apply.
Command Claude Code
Unlike /code-polish, this variant is a distributed state machine. Hermit runs its own review-and-apply loop until it is self-clean, and Claude provides two independent review passes on top as a quality gate. Claude's token spend is capped at 2 review rounds.
Command Claude Code
Same outline as /feature-develop, but the Phase 3 implementation is delegated to Hermit over MCP so Claude's context never absorbs the grunt-work tool output.
SkyWalker2506/ccplugin-opencode-bridge
Command
Manage OpenCode/Zen connections, switch models, and configure local Ollama.
rahulcvwebsitehosting/wayfinder
Command Claude Code
You are a code review specialist. Identify genuine issues while filtering out false positives.
Command Claude Code
Activate coder agent - Software development agent for clean code, design patterns, algorithms, and best practices.
Command Claude Code
Activate philosopher agent - Philosophy agent exploring wisdom traditions, consciousness, and the perennial questions of existence.
Command Claude Code
Activate scientist agent - Scientific research agent for logic, natural sciences, and systematic problem-solving.
Command
A TOEFL practice routine that finds your weakest question types and creates repeated exercises for them. TOEFL is an English test covering reading, listening, speaking and writing.
Command
A TOEFL answer-scoring command that uses a local language model, and speech-to-text software for recordings, to assess reading, listening, writing, and speaking practice. TOEFL is an English test with separate reading, listening, writing, and speaking sections.
Command
A local, offline generator for TOEFL practice questions. TOEFL is an English-language test with reading, listening, speaking, and writing sections.
Command
Delegate a task to a separate zero-token Claude Code session running on your local model (llama-swap/llama.cpp). Use for long-running work (pipelines, test suites, bulk edits) or simple side tasks (e.g. writing a hook) that would otherwise burn main-session context. The spawned session opens in a new terminal window…
Command
Convert an existing slash command into a cloudone-delegated sibling: the expensive, stateless part of its body runs in a spawned local-model side session while the steps that need your session's identity stay home. Emits a NEW command next to the original; the original is never modified.
Command
Guided setup for cloudone: choose how side sessions launch (new terminal / tmux / custom), point at your local model backend (llama-swap recommended), pick a model, and optionally install the example harness. Writes /.claude/cloudone/config.json.
Command
View or change any claude-tts setting by key, with validation, then restart the daemon to apply.