maxh8086/tokenfrugal
Offload routine coding to a local LLM (Ollama/OpenAI-compatible) via an MCP gateway with Docker tools, 280+ personas and a Neo4j task plan. Save Claude/Codex tokens.
ARCHETYPE
Skill Collector
Ten-plus skills loaded on demand. A procedural-knowledge library.
CONTEXT TAX · EVERY TURN
up to ~5.2k tokens
Claude Code loads MCP tools on demand, so the real cost is usually lower.
Moderate · median rig: 2.3k · breakdown
Copy this rig
# review before running: this installs third-party code
$ npx degit maxh8086/tokenfrugal/.claude ./rig-tokenfrugal # inspect, then merge into .claude/ 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 maxh8086/tokenfrugal/router/skills/code-l1 .claude/skills/code-l1 $ npx degit maxh8086/tokenfrugal/router/skills/code-l2 .claude/skills/code-l2 $ npx degit maxh8086/tokenfrugal/router/skills/code-l3 .claude/skills/code-l3 $ npx degit maxh8086/tokenfrugal/router/skills/code-l4 .claude/skills/code-l4 $ npx degit maxh8086/tokenfrugal/router/skills/code-l5 .claude/skills/code-l5 $ npx degit maxh8086/tokenfrugal/router/skills/code-l6 .claude/skills/code-l6 $ npx degit maxh8086/tokenfrugal/router/skills/code-review .claude/skills/code-review $ npx degit maxh8086/tokenfrugal/router/skills/db-admin .claude/skills/db-admin $ npx degit maxh8086/tokenfrugal/router/skills/db-dev .claude/skills/db-dev $ npx degit maxh8086/tokenfrugal/router/skills/docs-style .claude/skills/docs-style $ npx degit maxh8086/tokenfrugal/router/skills/planning .claude/skills/planning $ npx degit maxh8086/tokenfrugal/router/skills/role-analyzer .claude/skills/role-analyzer $ npx degit maxh8086/tokenfrugal/router/skills/role-browser .claude/skills/role-browser $ npx degit maxh8086/tokenfrugal/router/skills/role-builder .claude/skills/role-builder $ npx degit maxh8086/tokenfrugal/router/skills/role-data .claude/skills/role-data $ npx degit maxh8086/tokenfrugal/router/skills/role-debugger .claude/skills/role-debugger $ npx degit maxh8086/tokenfrugal/router/skills/role-designer .claude/skills/role-designer $ npx degit maxh8086/tokenfrugal/router/skills/role-devops .claude/skills/role-devops $ npx degit maxh8086/tokenfrugal/router/skills/role-finance .claude/skills/role-finance $ npx degit maxh8086/tokenfrugal/router/skills/role-github-actions-read .claude/skills/role-github-actions-read $ npx degit maxh8086/tokenfrugal/router/skills/role-github-read .claude/skills/role-github-read $ npx degit maxh8086/tokenfrugal/router/skills/role-research .claude/skills/role-research $ npx degit maxh8086/tokenfrugal/router/skills/role-tracker .claude/skills/role-tracker $ npx degit maxh8086/tokenfrugal/router/skills/security-audit .claude/skills/security-audit $ npx degit maxh8086/tokenfrugal/router/skills/sre-triage .claude/skills/sre-triage
$ curl -fsSL --create-dirs -o .claude/agents/tokenfrugal.md https://raw.githubusercontent.com/maxh8086/tokenfrugal/main/.claude/agents/tokenfrugal.md This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 16.2k rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(./.env.*)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(git push --force:*)",
"Read(**/*.pem)",
"Read(.env)",
"Read(~/.aws/**)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Bash(rm -rf *)",
"Bash(sudo *)",
"Bash(sudo:*)",
"Bash(rm -rf /)",
"Bash(git push -f:*)",
"Read(.env.*)",
"Bash(git reset --hard*)",
"Bash(rm -rf /*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(gh pr merge *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(gh pr merge:*)",
"Bash(gh pr create *)",
"Bash(git reset --hard:*)",
"Bash(npm publish:*)",
"Bash(git rebase *)"
]
}
} MCP servers (1)
| server | source | est. tokens |
|---|---|---|
| tokenfrugal | local / custom | 2.5k |
Skills (25)
code-l1code-l2code-l3code-l4code-l5code-l6code-reviewdb-admindb-devdocs-styleplanningrole-analyzerrole-browserrole-builderrole-datarole-debuggerrole-designerrole-devopsrole-financerole-github-actions-readrole-github-readrole-researchrole-trackersecurity-auditsre-triage
Subagents (1)
| tokenfrugal model: haiku | Delegates routine work (code, tests, docstrings, lint fixes, log triage, summaries) to local models through the tokenfrugal MCP server and returns the capped summary. Use for scoped tasks with a clear |
Similar rigs
pravindurgani/claude-code-multipane-iterm2
Stop your AI agent from reviewing its own code — 4-pane iTerm2 setup with role-locked Claude Code sessions. Step-by-step guide included.
Minimalist 1.0k tok · · Q0
HartBrook/staghorn
Sync Claude Code configs and skills from GitHub — for teams, communities, or across your personal projects. Run evals to verify your guidelines actually work.
Minimalist 196 tok · · Q18
kimon1230/claude_extensions
Custom hooks, skills, and rules for Claude Code: adds automatic Python formatting, context persistence across sessions, parallel subagent workflows, and a status line.
Pragmatist 1.7k tok · · Q5
Jam0731/ParallaxAI
Multi-agent collaboration gateway for one-person companies. Route messages to specialized AI agents (Munger/Woz/Ogilvy/Taleb) with shared memory, session persistence, and cost tracking.
Pragmatist 3.3k tok · · Q0