SonyHarv/claude-polyrouter
Intelligent model routing for Claude Code — routes haiku/sonnet/opus by complexity. Live HUD with ctx%, rate limits & savings. 10 languages, zero config.
ARCHETYPE
Skill Collector
Ten-plus skills loaded on demand. A procedural-knowledge library.
Copy this rig
# review before running: this installs third-party code $ npx degit SonyHarv/claude-polyrouter/agents ./rig-claude-polyrouter/agents $ npx degit SonyHarv/claude-polyrouter/commands ./rig-claude-polyrouter/commands $ npx degit SonyHarv/claude-polyrouter/skills ./rig-claude-polyrouter/skills $ npx degit SonyHarv/claude-polyrouter/hooks ./rig-claude-polyrouter/hooks
MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ npx degit SonyHarv/claude-polyrouter/skills/config .claude/skills/config $ npx degit SonyHarv/claude-polyrouter/skills/dashboard .claude/skills/dashboard $ npx degit SonyHarv/claude-polyrouter/skills/knowledge .claude/skills/knowledge $ npx degit SonyHarv/claude-polyrouter/skills/learn-off .claude/skills/learn-off $ npx degit SonyHarv/claude-polyrouter/skills/learn-on .claude/skills/learn-on $ npx degit SonyHarv/claude-polyrouter/skills/learn-reset .claude/skills/learn-reset $ npx degit SonyHarv/claude-polyrouter/skills/learn .claude/skills/learn $ npx degit SonyHarv/claude-polyrouter/skills/retry .claude/skills/retry $ npx degit SonyHarv/claude-polyrouter/skills/route .claude/skills/route $ npx degit SonyHarv/claude-polyrouter/skills/stats .claude/skills/stats
$ curl -fsSL --create-dirs -o .claude/agents/deep-executor.md https://raw.githubusercontent.com/SonyHarv/claude-polyrouter/master/agents/deep-executor.md $ curl -fsSL --create-dirs -o .claude/agents/fast-executor.md https://raw.githubusercontent.com/SonyHarv/claude-polyrouter/master/agents/fast-executor.md $ curl -fsSL --create-dirs -o .claude/agents/opus-orchestrator.md https://raw.githubusercontent.com/SonyHarv/claude-polyrouter/master/agents/opus-orchestrator.md $ curl -fsSL --create-dirs -o .claude/agents/standard-executor.md https://raw.githubusercontent.com/SonyHarv/claude-polyrouter/master/agents/standard-executor.md
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 7,204 rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(rm -rf *)",
"Read(**/*.pem)",
"Bash(rm -rf /)",
"Bash(git push --force:*)",
"Bash(sudo *)",
"Read(.env)",
"Bash(rm -rf /*)",
"Read(./.env.*)",
"Read(~/.aws/**)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(sudo:*)",
"Bash(git reset --hard*)",
"Bash(git reset --hard:*)",
"Read(.env.*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(rm:*)",
"Bash(git rebase *)",
"Bash(wget *)",
"Bash(npm publish:*)",
"Bash(git commit *)",
"Bash(gh pr merge *)"
]
}
} Skills (10)
Subagents (4)
| deep-executor model: claude-opus-4-7 | Complex analysis using the deep-tier model |
| fast-executor model: claude-haiku-4-5 | Quick answers using the fast-tier model |
| opus-orchestrator model: claude-opus-4-7 | Orchestrates complex multi-step tasks, delegates subtasks to cheaper tiers |
| standard-executor model: claude-sonnet-4-6 | Standard coding tasks using the standard-tier model |
Slash commands (13)
/advisor/config/dashboard/effort/export/knowledge/learn-off/learn-on/learn-reset/learn/retry/route/stats
Similar rigs
kirilxd/claude-tutor
Turn Claude Code into your personal tutor — personalized learning plans, adaptive quizzes, SM-2 spaced repetition, and a web dashboard. Works with any topic.
Pragmatist 1.5k tok ·
ChristopherALewis/claude-et-ay
Estimated time to completion for Claude Code: a live status line ETA for the prompt Claude is working on, learnt from your own history.
Pragmatist 183 tok ·
Luke-Bradford/autonomy-engine
Repo-agnostic engine for running Claude Code autonomy loops against any target repo
Pragmatist 9.9k tok ·
tigerless-labs/autoharness
Autoharness — a self-learning skill layer for Claude Code — distills skills from your real sessions, updates them as you work, and prunes the ones that stop getting used. No daemon, no benchmark.
Pragmatist 2.6k tok ·