{"_id":"@elixpo/claudeops","_rev":"3-b7d437e2989ea80a15213468857fc827","name":"@elixpo/claudeops","dist-tags":{"latest":"1.0.3"},"versions":{"1.0.0":{"name":"@elixpo/claudeops","version":"1.0.0","keywords":["claude","claude-code","ai","agents","mcp","optimization","token-savings","devtools"],"author":{"name":"Ayushman Bhattacharya","email":"ayushman@myceli.ai"},"license":"MIT","_id":"@elixpo/claudeops@1.0.0","maintainers":[{"name":"elixpo","email":"ayushbhatt633@gmail.com"}],"homepage":"https://github.com/elixpo/claudeops","bugs":{"url":"https://github.com/elixpo/claudeops/issues"},"bin":{"claudeops":"cli.js"},"dist":{"shasum":"2fba3cf3b05f548456839fe44721bef8c53b8f56","tarball":"https://registry.npmjs.org/@elixpo/claudeops/-/claudeops-1.0.0.tgz","fileCount":36,"integrity":"sha512-x3LuB4gtDgi4aLXopwDPFZKrunFDVCRyG6qOgclz4nxhLu2IOKFXOMc03EC3sVDVkdqykj5LdFeJmnYiBBnVqg==","signatures":[{"sig":"MEYCIQCatsEpqfsUBu3eHU5XfoKnLrrUmwJkP8siaEIljIClZgIhAKxq3xzp8h5sft7xdD+3cdqmIXmZC0+UfFLE8zI9Gmyv","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"unpackedSize":122347},"engines":{"node":">=20"},"gitHead":"aa3e1c5a4c6a8b0614ea2afb12c404e8c61c0b06","_npmUser":{"name":"elixpo","email":"ayushbhatt633@gmail.com"},"repository":{"url":"git+https://github.com/elixpo/claudeops.git","type":"git"},"_npmVersion":"11.7.0","description":"Drop-in optimization toolkit for Claude Code — 15 agents, token savings, safety hooks, UI pipeline","directories":{},"_nodeVersion":"20.19.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/claudeops_1.0.0_1776333235348_0.931224537921659","host":"s3://npm-registry-packages-npm-production"}},"1.0.2":{"name":"@elixpo/claudeops","version":"1.0.2","keywords":["claude","claude-code","claude-agents","claude-mcp","claude-hooks","ai-agents","ai-coding","ai-tools","mcp-server","mcp","optimization","token-savings","code-review","devtools","developer-tools","cli","automation","llm","agentic","vibe-coding"],"author":{"name":"Ayushman Bhattacharya","email":"ayushman@myceli.ai"},"license":"MIT","_id":"@elixpo/claudeops@1.0.2","maintainers":[{"name":"elixpo","email":"ayushbhatt633@gmail.com"}],"homepage":"https://github.com/elixpo/claudeOps.elixpo","bugs":{"url":"https://github.com/elixpo/claudeOps.elixpo/issues"},"bin":{"claudeops":"cli.js"},"dist":{"shasum":"7f1c2a0feda7bc807d883e2dcc68228bd02cde10","tarball":"https://registry.npmjs.org/@elixpo/claudeops/-/claudeops-1.0.2.tgz","fileCount":36,"integrity":"sha512-IUWASs8md/7OICfzYLlms3q8LFj1k67ZMeTgVShs1nljx6sx1iXaMAU7m+At3CbTPl+MV1IkRymoF87iuMBoeA==","signatures":[{"sig":"MEUCICCOtG/8zz500R6eDNm81/5L9IDa7GKE/a6ilHwYRcICAiEAtErL21Iz1F36N5OvLGHO2T8E4ZB+xCNSTdupzkqs7Pg=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"unpackedSize":122935},"engines":{"node":">=20"},"gitHead":"68c381dc974d0f96485bc6edac49c388a550ec29","_npmUser":{"name":"elixpo","email":"ayushbhatt633@gmail.com"},"repository":{"url":"git+https://github.com/elixpo/claudeOps.elixpo.git","type":"git"},"_npmVersion":"11.7.0","description":"Drop-in optimization toolkit for Claude Code — 15 agents, token savings, safety hooks, UI pipeline","directories":{},"_nodeVersion":"20.19.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/claudeops_1.0.2_1776336221568_0.708657269232823","host":"s3://npm-registry-packages-npm-production"}},"1.0.3":{"name":"@elixpo/claudeops","version":"1.0.3","description":"Drop-in optimization toolkit for Claude Code — 15 agents, token savings, safety hooks, UI pipeline","bin":{"claudeops":"cli.js"},"keywords":["claude","claude-code","claude-agents","claude-mcp","claude-hooks","ai-agents","ai-coding","ai-tools","mcp-server","mcp","optimization","token-savings","code-review","devtools","developer-tools","cli","automation","llm","agentic","vibe-coding"],"author":{"name":"Ayushman Bhattacharya","email":"ayushman@myceli.ai"},"license":"MIT","repository":{"type":"git","url":"git+https://github.com/elixpo/claudeOps.elixpo.git"},"homepage":"https://github.com/elixpo/claudeOps.elixpo","engines":{"node":">=20"},"gitHead":"17baee85882680913cfd189aafc7ba33d442faef","_id":"@elixpo/claudeops@1.0.3","bugs":{"url":"https://github.com/elixpo/claudeOps.elixpo/issues"},"_nodeVersion":"20.19.2","_npmVersion":"11.7.0","dist":{"integrity":"sha512-owtDEnyQUauw1gZCS6N0vM7/Nhk5GUtAUzYbIzgD/EiohVCPNY4UVLuTVVRNfHcHFG1vgs8qrsgjq3JkTVE/Ag==","shasum":"ad8ecad13f2ff9657606e4546477fea089201f02","tarball":"https://registry.npmjs.org/@elixpo/claudeops/-/claudeops-1.0.3.tgz","fileCount":36,"unpackedSize":126245,"signatures":[{"keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U","sig":"MEUCIQD4uBya6UeKAmMfiLj/5tJVHRq05k+8rz4GGSBaRfKv4wIgQWkzp/jFrYAxO9aQlQ/XC7Mduocgz0W78p04xh2/BUc="}]},"_npmUser":{"name":"elixpo","email":"ayushbhatt633@gmail.com"},"directories":{},"maintainers":[{"name":"elixpo","email":"ayushbhatt633@gmail.com"}],"_npmOperationalInternal":{"host":"s3://npm-registry-packages-npm-production","tmp":"tmp/claudeops_1.0.3_1776607580684_0.04827793283960391"},"_hasShrinkwrap":false}},"time":{"created":"2026-04-16T09:53:55.251Z","modified":"2026-04-19T14:06:20.928Z","1.0.0":"2026-04-16T09:53:55.477Z","1.0.2":"2026-04-16T10:43:41.698Z","1.0.3":"2026-04-19T14:06:20.831Z"},"bugs":{"url":"https://github.com/elixpo/claudeOps.elixpo/issues"},"author":{"name":"Ayushman Bhattacharya","email":"ayushman@myceli.ai"},"license":"MIT","homepage":"https://github.com/elixpo/claudeOps.elixpo","keywords":["claude","claude-code","claude-agents","claude-mcp","claude-hooks","ai-agents","ai-coding","ai-tools","mcp-server","mcp","optimization","token-savings","code-review","devtools","developer-tools","cli","automation","llm","agentic","vibe-coding"],"repository":{"type":"git","url":"git+https://github.com/elixpo/claudeOps.elixpo.git"},"description":"Drop-in optimization toolkit for Claude Code — 15 agents, token savings, safety hooks, UI pipeline","maintainers":[{"name":"elixpo","email":"ayushbhatt633@gmail.com"}],"readme":"

\n \"ClaudeOps\"\n

ClaudeOps

\n

\n Make Claude Code actually work like a senior engineer.
\n Drop-in agents, automatic token compression, safety hooks, and a UI design pipeline — one command to install.\n

\n

\n \"npm\"\n \"Downloads\"\n \"Stars\"\n \"License\"\n \"Platform\"\n

\n

\n\n---\n\n## The Problem\n\nClaude Code is powerful, but out of the box it burns through tokens reading things it doesn't need, skips code review, writes orphan files nobody calls, and produces generic-looking UIs. You end up babysitting it.\n\nClaudeOps fixes that. It's a set of config files, agents, and hooks that drop into your Claude Code setup and make it work the way you'd expect a senior engineer to work — research before coding, review before merging, prove things work before saying \"done.\"\n\n---\n\n

\n \"ClaudeOps\"\n

\n\n--- \n\n\n## Quick Start\n\n```bash\nnpx @elixpo/claudeops init\n```\n\nThat's it. The installer walks you through each component — nothing gets installed without your say-so.\n\n**Other commands:**\n```bash\nnpx @elixpo/claudeops status # see what's installed\nnpx @elixpo/claudeops remove # clean uninstall\n```\n\n**Flags:**\n```bash\nnpx @elixpo/claudeops init --yes # skip all prompts, install everything\nnpx @elixpo/claudeops init --agents-only # only install the 15 agents\nnpx @elixpo/claudeops init --no-mcp # skip MCP server setup\n```\n\n---\n\n## What's Inside\n\n### Agents\n\n15 specialized agents that Claude auto-delegates to based on what you're doing. They live in `~/.claude/agents/` and cost zero tokens when they're not active.\n\n**Security & Review**\n| Agent | What it does |\n|-------|-------------|\n| `breaker` | Tries to break your code. Every issue comes with a concrete exploit, not a vague warning. |\n| `red-team` | You tell it what security measures you added. It attacks those specific assumptions. |\n| `refiner` | Multi-pass review: write, critique against 6 principles, rewrite, verify. Two modes — quick surgical or full systematic. |\n| `prism` | 4 isolated review passes (security, performance, test coverage, correctness), each with blinders so nothing gets glossed over. |\n| `test-auditor` | Catches fake tests — mocks that test nothing, assertions that always pass, happy-path-only coverage. |\n\n**Planning & Design**\n| Agent | What it does |\n|-------|-------------|\n| `architect` | System design and ADRs. Compares 3+ approaches before recommending. Never writes implementation code. |\n| `brancher` | Forces exploration of multiple solutions before committing. Won't start coding until all branches are compared. |\n| `questioner` | 6 mandatory questions before any code. Surfaces hidden assumptions and finds the simplest path. |\n| `specwriter` | Interviews you through 5 rounds of questions, then produces a SPEC.md detailed enough for a fresh session to execute without follow-ups. |\n| `judge` | Evaluates competing solutions through binary gates and weighted scoring. Picks a winner, never declares a tie. |\n\n**Research & Integration**\n| Agent | What it does |\n|-------|-------------|\n| `researcher` | 7 depth levels from `surface` (30 seconds) to `overkill` (unlimited). Launches parallel sub-agents at higher levels. |\n| `fuzzer` | Property-based testing across 1000+ random inputs. Finds edge cases humans miss. |\n| `wirer` | Verifies new code is actually wired into the running system. Catches the #1 AI coding failure: beautiful code that nothing calls. |\n\n**UI & Frontend**\n| Agent | What it does |\n|-------|-------------|\n| `ui-architect` | Auto-triggers on frontend tasks. Pulls from component libraries (shadcn, magicui, animotion) instead of writing from scratch. Handles responsive, dark mode, accessibility. |\n| `design-critic` | Auto-runs after UI implementation. Scores visual quality 1-10 with specific fix suggestions. |\n\n### Token Savings\n\nThese tools run automatically after setup — they intercept Claude's input and output and compress what doesn't need to be verbose.\n\n| Tool | What it does | Savings |\n|------|-------------|---------|\n| [RTK](https://github.com/rtk-ai/rtk) by [rtk-ai](https://github.com/rtk-ai) | Compresses bash command output, auto-rewrites `git`, `ls`, `cat`, test runners, and 100+ more | 60-90% |\n| [jCodeMunch](https://github.com/jgravelle/jcodemunch-mcp) | Lightweight code indexing via tree-sitter | 95%+ |\n| [Context Mode](https://github.com/mksglu/context-mode) | Sandboxes tool outputs in SQLite instead of dumping them into context | 98% |\n| [MCP Compressor](https://github.com/atlassian-labs/mcp-compressor) | Compresses MCP server responses | 70-95% |\n| [Cozempic](https://github.com/Ruya-AI/cozempic) | Auto-prunes stale context at multiple thresholds | 30-70% |\n| [Headroom](https://github.com/chopratejas/headroom) | AST-aware prompt compression | ~50% |\n| [Serena](https://github.com/oraios/serena) | LSP-powered code navigation (40+ languages) | Significant |\n| [Docfork](https://github.com/docfork/docfork) | 9000+ library docs served on demand | Moderate |\n\nThe installer asks before adding each one. Everything is optional.\n\n### UI Component Libraries\n\nIf you do frontend work, ClaudeOps connects Claude to real component registries so it fetches production components instead of hallucinating them.\n\n| Library | What you get |\n|---------|-------------|\n| [21st.dev Magic](https://github.com/21st-dev/magic-mcp) | AI-generated production components |\n| [shadcn/ui](https://ui.shadcn.com/docs/registry/mcp) | Official registry, zero hallucinations |\n| [Magic UI](https://magicui.design/docs/mcp) | 60+ animated components |\n| [Animotion](https://animotion-mcp.github.io/) | 745 CSS animations, 9000 SVG icons |\n| [Aceternity UI](https://github.com/rudra016/aceternityui-mcp) | 200+ cinematic and 3D components |\n| [Glance](https://github.com/DebugBase/glance) | Browser screenshots — Claude sees what it built |\n\n> 21st.dev needs a free API key from [21st.dev](https://21st.dev). The installer reminds you.\n\n### Safety Hooks\n\nThese run deterministically on every session — no LLM involved, just bash and regex.\n\n| What it catches | When |\n|----------------|------|\n| Secrets in prompts or files (AWS, GitHub, Stripe, SSH keys, JWTs, and more) | Before every prompt and file write |\n| Destructive commands (`rm -rf`, `git reset --hard`, `DROP TABLE`, `git clean -f`, etc.) | Before every bash command |\n| `--no-verify` on git commits | Before git commits |\n| Commits to main/master | Before git commits (warning, not blocking) |\n| Orphan files with zero import references | On file edits and session end |\n| Stuck loops (3+ identical failures) | After bash commands |\n| Session continuity | Writes HANDOVER.md on exit, reads it on next session start |\n\n### Per-Project Setup\n\nAfter the global install, each project just needs a `.claudeignore` file (a template is included) to keep Claude from reading `node_modules/`, `dist/`, and other noise. Code indexing happens automatically on first visit.\n\n---\n\n## How It Works\n\nThe approach is simple: use Claude Code's own extension points (agents, hooks, MCP servers, plugins) to make Claude behave better by default.\n\n- **Agents over instructions** — a specialized agent that loads on-demand costs zero tokens when unused. A rule in CLAUDE.md costs tokens every single turn.\n- **Hooks over suggestions** — a PreToolUse hook that blocks a destructive command is stronger than a polite instruction Claude might ignore.\n- **Evidence over assertions** — every verification requires proof (test output, type checker results), not \"this should work.\"\n- **Autonomous over manual** — tools that intercept and optimize automatically beat workflows that depend on you remembering to invoke them.\n\n---\n\n## Uninstall\n\n```bash\nnpx @elixpo/claudeops remove\n```\n\nInteractively removes agents, hooks, MCP servers, and plugins. Your non-ClaudeOps configuration is preserved.\n\n---\n\n## Credits\n\nClaudeOps bundles and configures tools built by the wider Claude Code community. Full credit to their authors — ClaudeOps just wires them into a single install flow.\n\n- **[RTK](https://github.com/rtk-ai/rtk)** by [rtk-ai](https://github.com/rtk-ai) — the CLI output compression layer that powers 60-90% of the bash token savings. Installed via the official installer (or Homebrew / cargo) and wired into Claude Code's PreToolUse hook via `rtk init -g` during `claudeops init`.\n- [jCodeMunch](https://github.com/jgravelle/jcodemunch-mcp), [Context Mode](https://github.com/mksglu/context-mode), [MCP Compressor](https://github.com/atlassian-labs/mcp-compressor), [Cozempic](https://github.com/Ruya-AI/cozempic), [Headroom](https://github.com/chopratejas/headroom), [Serena](https://github.com/oraios/serena), [Docfork](https://github.com/docfork/docfork) — token-saving MCPs and tools integrated into the install pipeline.\n- UI registries: [21st.dev](https://21st.dev), [shadcn/ui](https://ui.shadcn.com), [Magic UI](https://magicui.design), [Animotion](https://animotion-mcp.github.io/), [Aceternity UI](https://github.com/rudra016/aceternityui-mcp), [Glance](https://github.com/DebugBase/glance).\n\n## Contributing\n\nPRs welcome. Areas where help is needed:\n\n- New agents for specific domains (ML, mobile, embedded)\n- Benchmarks comparing with/without ClaudeOps\n- Platform-specific install improvements\n\n
\n\n--- \n\n

\n Built with ❤️ and maintained by @elixpo\n

\n\n\n\n \n \n \"github-snake\"\n\n\n\n","readmeFilename":"README.md"}