[{"data":1,"prerenderedAt":4},["ShallowReactive",2],{"readme:maggy":3},"\u003Ch1>Claude Bootstrap + Maggy\u003C\u002Fh1>\n\u003Cblockquote>\n\u003Cp>\u003Cstrong>Turn Claude Code into a self-reviewing, test-enforced engineering system that remembers context across sessions — then route work across 13 models from a single dashboard.\u003C\u002Fstrong>\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Cp>Claude Bootstrap is an installable config pack (skills, hooks, rules, templates) for Claude Code. Maggy is the optional local server that adds multi-model routing, a web dashboard, intent-driven protocols, and plugin orchestration. Both live in this repo. Start with Bootstrap; add Maggy when you need the harness.\u003C\u002Fp>\n\u003Cp>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Falinaqi\u002Fmaggy\u002Fblob\u002FHEAD\u002Fmaggy\u002Ftests\u002F\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Ftests-1100%2B%20passing-brightgreen\" alt=\"Tests\" \u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Falinaqi\u002Fmaggy\u002Fblob\u002FHEAD\u002FCHANGELOG.md\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fversion-6.37.0-blue\" alt=\"Version\" \u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Falinaqi\u002Fmaggy\u002Fstargazers\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Falinaqi\u002Fmaggy\" alt=\"Stars\" \u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Falinaqi\u002Fmaggy\u002Fblob\u002FHEAD\u002FLICENSE\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Flicense-MIT-green\" alt=\"License: MIT\" \u002F>\u003C\u002Fa>\u003C\u002Fp>\n\u003Cp>\u003Cstrong>1100+ tests. 67 skills. 15 MCP tools. Used daily across production codebases.\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Chr \u002F>\n\u003Ch2>Who This Is For\u003C\u002Fh2>\n\u003Cul>\n\u003Cli>\u003Cstrong>Solo engineers\u003C\u002Fstrong> using Claude Code who want TDD enforcement, quality gates, and memory that survives context compaction — without changing their workflow\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Teams\u003C\u002Fstrong> routing work across Claude, DeepSeek, Kimi, Gemini, and Codex from a single dashboard with cost-aware model selection\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Platform engineers\u003C\u002Fstrong> building AI-assisted developer tooling who need a reference implementation with intent tracking, protocol execution, and plugin architecture\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Chr \u002F>\n\u003Ch2>Choose Your Path\u003C\u002Fh2>\n\u003Ctable>\n\u003Cthead>\n\u003Ctr>\n\u003Cth>\u003C\u002Fth>\n\u003Cth>Claude Bootstrap\u003C\u002Fth>\n\u003Cth>Maggy Harness\u003C\u002Fth>\n\u003C\u002Ftr>\n\u003C\u002Fthead>\n\u003Ctbody>\u003Ctr>\n\u003Ctd>\u003Cstrong>What it is\u003C\u002Fstrong>\u003C\u002Ftd>\n\u003Ctd>Skills, hooks, rules installed into \u003Ccode>~\u002F.claude\u002F\u003C\u002Fcode>\u003C\u002Ftd>\n\u003Ctd>Local FastAPI server + web dashboard\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>\u003Cstrong>Install time\u003C\u002Fstrong>\u003C\u002Ftd>\n\u003Ctd>~30 seconds\u003C\u002Ftd>\n\u003Ctd>~5 minutes (Python 3.11+, API keys)\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>\u003Cstrong>Requires\u003C\u002Fstrong>\u003C\u002Ftd>\n\u003Ctd>Claude Code (also works with Codex, Kimi, Gemini CLI)\u003C\u002Ftd>\n\u003Ctd>Everything in Bootstrap + Python + optional Docker\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>\u003Cstrong>You get\u003C\u002Fstrong>\u003C\u002Ftd>\n\u003Ctd>TDD enforcement, 67 skills, quality gates, ADR reviews, iCPG, Mnemos memory\u003C\u002Ftd>\n\u003Ctd>All of Bootstrap + 13-tier routing, skill protocols, Telos testing, Cortex MCP, plugins, dashboard\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftbody>\u003C\u002Ftable>\n\u003Ch3>Bootstrap — 30-second install\u003C\u002Fh3>\n\u003Cpre>\u003Ccode class=\"language-bash\">git clone https:\u002F\u002Fgithub.com\u002Falinaqi\u002Fmaggy.git\ncd maggy && .\u002Finstall.sh\n\u003C\u002Fcode>\u003C\u002Fpre>\n\u003Cp>Your next Claude Code session picks it up automatically.\u003C\u002Fp>\n\u003Ch3>Full Harness — zero-config\u003C\u002Fh3>\n\u003Cpre>\u003Ccode class=\"language-bash\">pipx install maggy-harness   # or: pip install maggy-harness\nmaggy bootstrap              # installs skills, hooks, ~\u002Fbin model wrappers, plugins\nmaggy serve                  # auto-configures from your local repos,\n                             # then opens the dashboard at localhost:8080\n\u003C\u002Fcode>\u003C\u002Fpre>\n\u003Cp>(or from source: \u003Ccode>cd maggy && .\u002Finstall.sh && maggy serve\u003C\u002Fcode>)\u003C\u002Fp>\n\u003Cp>No API keys required to start — Maggy runs in local mode and, on first launch,\ndiscovers your local git repos and opens the dashboard pointed at them. Add\n\u003Ccode>GITHUB_TOKEN\u003C\u002Fcode> \u002F \u003Ccode>ANTHROPIC_API_KEY\u003C\u002Fcode> later only if you want GitHub sync or\nAPI-model features. See \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Falinaqi\u002Fmaggy\u002Fblob\u002FHEAD\u002FGETTING_STARTED.md\" rel=\"nofollow ugc noopener\">GETTING_STARTED.md\u003C\u002Fa> for details.\u003C\u002Fp>\n\u003Chr \u002F>\n\u003Ch2>What It Looks Like in Practice\u003C\u002Fh2>\n\u003Cp>\u003Cstrong>Routing a task:\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Cpre>\u003Ccode>You: \"review the auth middleware for timing attacks\"\n→ Blast score: 8\u002F10 (security + architecture)\n→ Routed to: Claude (Tier 11)\n→ ADR gate: found docs\u002Fadr\u002F0003-jwt-strategy.md → injected as context\n→ Review runs with full architectural context\n\u003C\u002Fcode>\u003C\u002Fpre>\n\u003Cp>\u003Cstrong>Skill Protocol execution:\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Cpre>\u003Ccode>You: \"push to git\"\n→ Intent matched: git-push protocol\n→ ✅ lint       (2.1s)\n→ ✅ typecheck   (4.3s)\n→ ✅ tests       (11.2s)\n→ ✅ stage\n→ ✅ commit      [AI-generated: \"fix: resolve token refresh race condition\"]\n→ ✅ push\n\u003C\u002Fcode>\u003C\u002Fpre>\n\u003Cp>\u003Cstrong>Fatigue-aware memory:\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Cpre>\u003Ccode>Session fatigue: 0.61 (PRE-SLEEP)\n→ Mnemos: auto-checkpoint written\n→ Micro-consolidation: 3 ResultNodes compressed\n→ iCPG context injected: 2 ReasonNodes, 1 constraint\n→ Context freed: ~18k tokens\n\u003C\u002Fcode>\u003C\u002Fpre>\n\u003Chr \u002F>\n\u003Ch2>The Problem This Solves\u003C\u002Fh2>\n\u003Cp>You're using Claude Code. It's impressive — but:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>It picks the most expensive model for everything, including trivial tasks\u003C\u002Fli>\n\u003Cli>Context fills up, state is lost, you re-explain yourself every session\u003C\u002Fli>\n\u003Cli>There's no enforcement: code quality, test coverage, and A\u003C\u002Fli>\n\u003C\u002Ful>\n",1785276849930]